1. No category

1D/2D modelling suite for integral water solutions
DR
AF
T
SOBEK Suite
D-Flow 1D in Delta Shell
User Manual
DR
AF
T
T
DR
AF
SOBEK 3
D-Flow 1D in Delta Shell
User Manual
Version: 3.3.0
Revision: 37731
25 December 2014
DR
AF
T
SOBEK 3, User Manual
Published and printed by:
Deltares
Boussinesqweg 1
2629 HV Delft
P.O. 177
2600 MH Delft
The Netherlands
For sales contact:
telephone: +31 88 335 81 88
fax:
+31 88 335 81 11
e-mail:
[email protected]
www:
http://www.deltaressystems.nl
telephone:
fax:
e-mail:
www:
+31 88 335 82 73
+31 88 335 85 82
[email protected]
http://www.deltares.nl
For support contact:
telephone: +31 88 335 81 00
fax:
+31 88 335 81 11
e-mail:
[email protected]
www:
http://www.deltaressystems.nl
Copyright © 2014 Deltares
All rights reserved. No part of this document may be reproduced in any form by print, photo
print, photo copy, microfilm or any other means, without written permission from the publisher:
Deltares.
Contents
Contents
1 Preface
1.1 About Delta Shell framework and the manual
1.2 About Deltares . . . . . . . . . . . . . . . .
1.3 Typographical conventions . . . . . . . . . .
1.4 Support . . . . . . . . . . . . . . . . . . . .
1.5 System requirements . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2 Module D-Flow 1D: Overview
5
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4 Module D-Flow 1D: All about the modeling process
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 Import . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1 Import modeldata on <Project> level . . . . . . . . . . .
4.2.2 Import a network from another model on <network> level
4.2.3 Import a network from GIS . . . . . . . . . . . . . . . . .
4.2.3.1 The GIS import wizard . . . . . . . . . . . . .
4.2.3.2 Import from personal geodatabase . . . . . . .
4.2.3.3 Import of culvert (profile) data . . . . . . . . . .
4.2.4 Import cross section profiles from <csv> . . . . . . . . .
4.2.5 Import time series from <csv> . . . . . . . . . . . . . .
4.3 Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.1 Setting up a network from scratch . . . . . . . . . . . . .
4.3.2 Nodes and branches . . . . . . . . . . . . . . . . . . . .
4.3.2.1 Nodes . . . . . . . . . . . . . . . . . . . . . .
4.3.2.2 Branches . . . . . . . . . . . . . . . . . . . .
4.3.2.3 Interpolation across nodes . . . . . . . . . . .
4.3.3 Weir . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.3.1 Introduction . . . . . . . . . . . . . . . . . . .
4.3.3.2 Simple weir . . . . . . . . . . . . . . . . . . .
4.3.3.3 Gated weir . . . . . . . . . . . . . . . . . . . .
4.3.3.4 Weir with piers . . . . . . . . . . . . . . . . . .
4.3.3.5 Weir with detailed description of crest . . . . .
4.3.3.6 Free form weir . . . . . . . . . . . . . . . . . .
4.3.3.7 General structure . . . . . . . . . . . . . . . .
4.3.4 Pump . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Deltares
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
T
.
.
.
.
.
.
.
.
.
.
.
.
DR
AF
3 Module D-Flow 1D: Getting started
3.1 Introduction . . . . . . . . . . .
3.2 Starting a D-flow 1D model . . .
3.3 Schematization . . . . . . . . .
3.4 Generating a computational grid
3.5 Boundary conditions . . . . . .
3.6 Roughness . . . . . . . . . . .
3.7 Initial conditions . . . . . . . . .
3.8 Model parameter settings . . . .
3.9 Set output . . . . . . . . . . . .
3.10 Validation . . . . . . . . . . . .
3.11 Running a simulation . . . . . .
3.12 Viewing simulation results . . .
1
1
1
2
3
3
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
7
7
7
11
12
13
14
15
16
17
18
18
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
21
21
21
21
22
23
23
26
26
27
28
30
30
31
31
32
33
34
34
35
36
37
38
39
40
41
iii
SOBEK 3, User Manual
Culvert, Syphon and Inverted Syphon . . . . . . . . . . . . . . . . .
Composite structure . . . . . . . . . . . . . . . . . . . . . . . . . .
Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Extra Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Lateral Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Retention area . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Observation point . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cross Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.12.1 Adding Cross Sections to the network . . . . . . . . . . .
4.3.12.2 Cross Section YZ . . . . . . . . . . . . . . . . . . . . . .
4.3.12.3 Cross Section XYZ . . . . . . . . . . . . . . . . . . . . .
4.3.12.4 Cross Section ZW . . . . . . . . . . . . . . . . . . . . . .
4.3.12.5 Cross Section . . . . . . . . . . . . . . . . . . . . . . . .
4.3.12.6 Working with Shared Cross Section definitions . . . . . . .
4.3.12.7 Import and export cross sections from/to <csv>-file . . . .
4.3.12.8 Inspect multiple cross sections in one view . . . . . . . . .
4.3.13 General functions on network objects . . . . . . . . . . . . . . . . .
4.3.13.1 Esc key . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.13.2 Copy and paste network object . . . . . . . . . . . . . . .
4.3.13.3 Add network object . . . . . . . . . . . . . . . . . . . . .
4.3.13.4 Zoom to network object . . . . . . . . . . . . . . . . . . .
4.3.13.5 Selection of multiple network objects . . . . . . . . . . . .
4.4 Boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.1 Types of boundary conditions . . . . . . . . . . . . . . . . . . . . .
4.4.2 Editing boundary conditions . . . . . . . . . . . . . . . . . . . . . .
4.4.3 Time series for boundary conditions . . . . . . . . . . . . . . . . . .
4.4.4 Remarks on discharge boundary conditions in D-Flow 1D . . . . . .
4.4.4.1 Simulation results corresponding to discharge boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.4.2 Discharge-waterlevel-relation . . . . . . . . . . . . . . . .
4.5 Initial conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5.1 Setting the initial conditions . . . . . . . . . . . . . . . . . . . . . .
4.5.2 Initial conditions from restart . . . . . . . . . . . . . . . . . . . . . .
4.6 Roughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6.2 Defining roughness . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6.3 Import and export roughness from/to csv-file . . . . . . . . . . . . .
4.7 Wind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.8 Salt water intrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.9 Computational grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10 Model properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.2 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.3 Initial conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.4 Model settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.4.1 Roughness for tidal flow . . . . . . . . . . . . . . . . . . .
4.10.4.2 Salt water intrusion . . . . . . . . . . . . . . . . . . . . .
4.10.5 Output parameters . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.6 Run parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10.6.1 Simulation period and timestep . . . . . . . . . . . . . . .
DR
AF
T
4.3.5
4.3.6
4.3.7
4.3.8
4.3.9
4.3.10
4.3.11
4.3.12
iv
42
44
45
46
46
48
48
48
48
49
50
51
52
52
53
54
55
55
55
55
55
55
57
57
58
59
60
60
61
62
62
63
65
65
65
69
69
71
75
78
78
78
78
78
78
79
79
79
79
Deltares
Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
80
80
81
81
82
82
82
83
83
83
84
84
85
88
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
89
90
91
93
94
94
94
95
96
96
.
.
.
.
.
.
99
99
100
100
100
101
101
T
4.10.6.2 Restart and save State . . . . . . . . . . .
4.10.6.3 Model parameters . . . . . . . . . . . . . .
4.10.6.4 Structure Inertia Damping Factor . . . . . .
4.10.6.5 Quasi steady-state . . . . . . . . . . . . .
4.10.6.6 Extra resistance for general structure . . . .
4.10.6.7 Summerdike . . . . . . . . . . . . . . . . .
4.10.6.8 Advanced options . . . . . . . . . . . . . .
4.10.6.9 Volumes based on waterlevels or discharges
4.10.6.10 Reduction of timestep on large lateral flow .
4.10.6.11 Use timestep reduction on structure . . . .
4.10.6.12 Parameter set for lowland rivers . . . . . .
4.10.7 Default bed roughness . . . . . . . . . . . . . . . . .
4.11 Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.12 Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
DR
AF
5 Module D-Flow 1D: Simulation and model output
5.1 Simulation information . . . . . . . . . . . .
5.2 Results in the Map . . . . . . . . . . . . . .
5.3 Results in a Graph . . . . . . . . . . . . . .
5.4 Results in a Table . . . . . . . . . . . . . . .
5.5 Sideviews . . . . . . . . . . . . . . . . . . .
5.5.1 Routes . . . . . . . . . . . . . . . .
5.5.2 Results in Sideview . . . . . . . . . .
5.6 Export . . . . . . . . . . . . . . . . . . . . .
5.7 Case analysis . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6 Module D-Flow 1D: Morphology and Sediment Transport
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2 Input files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3 Output files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4 Scripting support . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4.1 Generating input files and working with spatially varying input
6.4.2 Dumping and dredging . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
References
A How to use OpenDA for Delta Shell models
A.1 Introduction . . . . . . . . . . . . . . . . . . . . . .
A.2 The Stochastic Model configuration . . . . . . . . .
A.2.1 Configuration for calibration . . . . . . . . .
A.2.2 Configuration for Ensemble Kalman Filtering
A.3 The Model configuration . . . . . . . . . . . . . . .
A.4 Installing OpenDA for Delta Shell models . . . . . .
A.5 Running the OpenDA application . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
103
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
B Appendix: Morphology and Sediment Transport
B.1 Input files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.1.1 Sediment input file . . . . . . . . . . . . . . . . . . . . . .
B.1.2 Morphology input file . . . . . . . . . . . . . . . . . . . . .
B.1.3 Sediment transport input file . . . . . . . . . . . . . . . . .
B.1.4 Sediment transport and morphology boundary condition file
B.1.5 Nodal Relations Definition file . . . . . . . . . . . . . . . .
Deltares
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
105
105
105
105
106
107
110
110
.
.
.
.
.
.
111
111
111
113
115
118
119
v
SOBEK 3, User Manual
B.2
B.3
B.5
vi
DR
AF
T
B.4
B.1.6 Table file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Output files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bedload sediment transport of non-cohesive sediment . . . . . . . . . . . .
B.3.1 Basic formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.3.2 Calculation of bedload transport at open boundaries . . . . . . . . .
Transport formulations for non-cohesive sediment . . . . . . . . . . . . . . .
B.4.1 Van Rijn (1993) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.2 Engelund-Hansen (1967) . . . . . . . . . . . . . . . . . . . . . . .
B.4.3 Meyer-Peter-Muller (1948) . . . . . . . . . . . . . . . . . . . . . . .
B.4.4 General formula . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.5 Bijker (1971) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.5.1 Basic formulation . . . . . . . . . . . . . . . . . . . . . .
B.4.5.2 Transport in wave propagation direction (Bailard-approach)
B.4.6 Van Rijn (1984) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.7 Soulsby/Van Rijn . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.8 Soulsby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.9 Ashida–Michiue (1974) . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.10 Wilcock–Crowe (2003) . . . . . . . . . . . . . . . . . . . . . . . . .
B.4.11 Gaeuman et al. (2009) laboratory calibration . . . . . . . . . . . . .
B.4.12 Gaeuman et al. (2009) Trinity River calibration . . . . . . . . . . . .
Morphological updating . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
120
121
121
121
122
122
122
127
127
128
128
129
130
132
134
135
138
139
139
140
141
Deltares
List of Figures
List of Figures
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
Data Import window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Data import window for network (features) . . . . . . . . . . . . . . . . . . .
The GIS import wizard . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example of the mapping table . . . . . . . . . . . . . . . . . . . . . . . . .
Import properties window for snapping precision and saving of mapping files
Setting of related tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example importing YZ Cross Section from <csv>-file . . . . . . . . . . . .
Selecting delimiters for a csv file . . . . . . . . . . . . . . . . . . . . . . . .
Selecting the columns of the <csv>-file . . . . . . . . . . . . . . . . . . . .
Linking a Timeseries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example of boundary nodes . . . . . . . . . . . . . . . . . . . . . . . . . .
Two branches with different Order number: No interpolation across the connection node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Two branches with same Order numbers: Bed level is interpolated across the
connection node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simple weir editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Gated weir editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Weir with piers editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Weir with detailed description of crest editor, the side-view shows the shape of
the crest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Free form weir editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General structure editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pump editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Culvert editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example of a Composite Structure in the Central Map . . . . . . . . . . . . .
Region window with a Composite Structure consisting of two weirs, a pump
and a culvert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bridge editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Editor for lateral source data . . . . . . . . . . . . . . . . . . . . . . . . . .
Generate data series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cross Section editor for yz Cross Sections . . . . . . . . . . . . . . . . . .
Editing window for an XYZ Cross Section . . . . . . . . . . . . . . . . . . .
Projection of a xyz-cross- section . . . . . . . . . . . . . . . . . . . . . . . .
DR
AF
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
4.12
Map view with open street background map and a D-Flow 1D branch generated near the city of Rotterdam . . . . . . . . . . . . . . . . . . . . . . . . .
Example of a cross section . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example of a weir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Editor for lateral sources/sinks . . . . . . . . . . . . . . . . . . . . . . . . .
Example of the resulting schematization . . . . . . . . . . . . . . . . . . . .
Computational grid editor . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Boundary nodes in the Central Map . . . . . . . . . . . . . . . . . . . . . .
Constant water level boundary condition . . . . . . . . . . . . . . . . . . . .
Editing the roughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Output options in the Properties Window . . . . . . . . . . . . . . . . . . .
Output options in the Properties Window . . . . . . . . . . . . . . . . . . .
Map results of water level . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results of water level for three locations along the branch in Function view .
Chart and the corresponding Properties window . . . . . . . . . . . . . . .
T
3.1
4.13
4.14
4.15
4.16
4.17
4.18
4.19
4.20
4.21
4.22
4.23
4.24
4.25
4.26
4.27
4.28
4.29
Deltares
8
9
10
10
11
12
13
13
14
15
17
18
19
20
22
23
24
25
25
26
27
28
29
30
32
33
34
36
37
38
39
40
41
42
43
44
44
45
47
47
49
50
50
vii
SOBEK 3, User Manual
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10
6.1
viii
DR
AF
T
4.30 Cross section editor for ZW Cross Sections . . . . . . . . . . . . . . . . . .
4.31 Cross section editor for Trapezium . . . . . . . . . . . . . . . . . . . . . . .
4.32 Switch between Local Cross Section definition and Shared Cross Section definition in the Cross Section editing window . . . . . . . . . . . . . . . . . . .
4.33 Example importing YZ Cross Section from <csv>-file . . . . . . . . . . . .
4.34 Example of a network with nodes with or without boundary conditions . . . .
4.35 Boundary nodes in the Central Map . . . . . . . . . . . . . . . . . . . . . .
4.36 Timeseries on boundary node . . . . . . . . . . . . . . . . . . . . . . . . .
4.37 Computational grid of a simple network with a discharge boundary condition
upstream (water flows from right to left). . . . . . . . . . . . . . . . . . . . .
4.38 Side-view of computed waterlevels . . . . . . . . . . . . . . . . . . . . . . .
4.39 Initial conditions editing window . . . . . . . . . . . . . . . . . . . . . . . .
4.40 write restart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.41 output states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.42 use restart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.43 Roughness editor for a model of the Dutch part of the river Meuse . . . . . .
4.44 Setting of roughness-sections in the Region window . . . . . . . . . . . . .
4.45 Cross section editor for an XYZ Cross Section with three Sections . . . . . .
4.46 Function table for roughness as a function of discharge and the graphical representation of the table content . . . . . . . . . . . . . . . . . . . . . . . . .
4.47 Wind shielding (factors) presented in the Central Map and the table for editing
4.48 Addition of salt in a flow model in the Properties window . . . . . . . . . . .
4.49 Project window after setting Use salinity to “True” . . . . . . . . . . . . . . .
4.50 The use of Thatcher-Harleman dispersion formulation . . . . . . . . . . . . .
4.51 Boundary node editor for salinity . . . . . . . . . . . . . . . . . . . . . . . .
4.52 Generate Computational Grid window . . . . . . . . . . . . . . . . . . . .
4.53 Table and map view of the computational grid (note that only waterlevel points
are shown in this view) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.54 Set output in the Properties window . . . . . . . . . . . . . . . . . . . . . .
4.55 Validation Report: example . . . . . . . . . . . . . . . . . . . . . . . . . . .
Output in the Project window . . . . . . . . . . . . . . . . . . . . . . .
Map results of discharge . . . . . . . . . . . . . . . . . . . . . . . . .
Layer properties editor . . . . . . . . . . . . . . . . . . . . . . . . . .
Customised map . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Select parameter for graphical representation . . . . . . . . . . . . . .
Time results of water level for 3 lcoations along the branch . . . . . . .
Example of 3 network routes shown in the network with different colours
Example of the use of intermediate locations to specify routes . . . . .
Example of sideview with Time Navigator . . . . . . . . . . . . . . . .
Example of Case analysis . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
51
52
53
54
57
58
59
60
60
62
63
64
64
65
66
67
68
70
72
73
74
75
76
77
86
88
.
.
.
.
.
.
.
.
.
.
90
91
92
93
93
94
95
95
96
97
How to simulate morfology together with a D-Flow 1D simulation . . . . . . .
99
Deltares
List of Tables
List of Tables
Options for roughness types and default values . . . . . . . . . . . . . . . .
85
A.1
A.1
A.2
Description of XML tags . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Description of XML tags . . . . . . . . . . . . . . . . . . . . . . . . . . . .
OpenDA program arguments . . . . . . . . . . . . . . . . . . . . . . . . . .
108
109
110
B.1
B.2
B.3
B.4
B.5
B.6
B.7
B.8
Sediment input file with keywords . . . . . . . . . . . . . . . . . . . . . . . 111
Options for sediment diameter characteristics . . . . . . . . . . . . . . . . . 112
Morphological input file with keywords . . . . . . . . . . . . . . . . . . . . . 113
Additional transport relations . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Transport formula parameters . . . . . . . . . . . . . . . . . . . . . . . . . 116
Nodal relation file with keywords . . . . . . . . . . . . . . . . . . . . . . . . 120
Additional transport relations . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Overview of the coefficients used in the various regression models (Soulsby
et al., 1993) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Overview of the coefficients used in the various regression models, continued
(Soulsby et al., 1993) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
DR
AF
B.9
T
4.1
Deltares
ix
DR
AF
T
SOBEK 3, User Manual
x
Deltares
1 Preface
1.1
About Delta Shell framework and the manual
This is the manual for SOBEK 3. With it’s first release in 2012, it will be extended in the
following years to become a full-fledged replacement for SOBEK 2. It is based on the newly
developed Delta Shell framework.
T
The Delta Shell framework is an integrated modelling environment which provides a platform
which can be used to integrate various models, data and tools. It consists of a framework
with a user interface which supports several environmental models. The program is easy to
configure and quick to learn. The graphically oriented interface is designed for intuitive use.
The user can download information from a variety of standard data formats and GIS systems.
DR
AF
This manual starts with an overview of the Delta Shell framework, how to set up the program
and start a project. It continues with - for each of the models - a tutorial to get started, a
description of the workflow to set up a model and an overview of the user interface which can
serve as a reference.
Currently the following models are available: D-Flow 1D (open water), D-RealTime Control,
D-WaterQuality 1D and D-Rainfall Runoff.
1.2
About Deltares
In 2008, four renowned Dutch organisations decided to pool their knowledge and expertise.
WL | Delft Hydraulics, GeoDelft, TNO’s Subsurface and Groundwater unit and parts of Rijkswaterstaat (the Dutch Directorate-General for Public Works and Water Management) joined
together to set up Deltares, an independent institute for the development, dissemination and
application of knowledge concerning water, soil and subsurface. The result is an organisation well-equipped to address complex, integrated issues relating to water, soil, subsurface
management and spatial planning in deltas, coastal areas and river basins.
Our national and international clients include government authorities, policy makers and administrators responsible for the short- and long-term governance of bodies of water and related infrastructures. Our clients are also in the private sector: for example, among multilateral
agencies, consulting engineers, contractors, and in industry. They all have one thing in common: the need for solid, practical advice.
Deltares brings together a long-standing reputation for excellence in hydrology, hydraulics,
morphology, water quality and ecology. Construction and design matters related to offshore,
coasts, harbours, estuaries, rivers and canals, and industry — also our forte — are approached in a manner tuned to the specific requirements of the client. In addition, we operate at the policy level by delivering decision support and carrying out environmental impact
assessments in the above mentioned working areas.
We have a full range of experimental facilities and computer programs — most of which have
been developed and validated by our experts in residence. On the basis of a sound understanding of the processes involved, all water systems can be simulated by us, numerically,
experimentally, or through a combination of the two. At Deltares, transfer of technology and
know-how is an inherent part of our approach. This is done through a variety of training
Deltares
1
SOBEK 3, User Manual
courses and seminars, and on the job.
Demands of increased production and economic growth are frequently coming in direct — and
public — conflict with environmental concerns. Balancing the needs of one with those of the
other, not only for today, but also for the future, is often expressed in the phrase "sustainable
development".
Typographical conventions
Throughout this manual, the following conventions help you to distinguish between different
elements of text to help you learn about SOBEK 3.
Example
Description
Waves
Boundaries
Title of a window or sub-window.
Sub-windows are displayed in the Module window and
cannot be moved.
Windows can be moved independently from the Module window, such as the Visualisation Area window.
Save
Item from a menu, title of a push button or the name of
a user interface input field.
Upon selecting this item (click or in some cases double
click with the left mouse button on it) a related action
will be executed; in most cases it will result in displaying
some other (sub-)window.
In case of an input field you are supposed to enter input
data of the required format and in the required domain.
DR
AF
1.3
T
No simple matter, it makes our clients’ jobs more demanding. In turn, they demand more of
us, not only for construction- and design- related issues, but also for far-reaching policy and
management concerns. They expect and get from us optimal performance: a multidisciplinary,
scientifically rigorous approach linked to cost-conscious good business sense.
<\tutorial\wave\swan-curvi>
<siu.mdw>
Directory names, filenames, and path names are expressed between angle brackets, <>. For the Linux
and UNIX environment a forward slash (/) is used instead of the backward slash (\) for PCs.
“27 08 1999”
Data to be typed by you into the input fields are displayed between double quotes.
Selections of menu items, option boxes etc. are described as such: for instance ‘select Save and go to
the next window’.
delft3d-menu
Commands to be typed by you are given in the font
Courier New, 10 points.
User actions are indicated with this arrow.
2
Deltares
Preface
1.4
Example
Description
[m/s] [-]
Units are given between square brackets when used
next to the formulae. Leaving them out might result in
misinterpretation.
Support
You should have the following information ready:
T
If you have a question about SOBEK 3 or any of the plug-ins for which you cannot find the
answer in the manual, you can contact Support at Deltares.
the version number of SOBEK 3 and the plug-ins (visible in the upper-left corner of the
DR
AF
window after the program is started);
the type of hardware you are using, including network hardware if applicable;
the operating system you are using;
the exact wording of any message that appeared on your screen (write it down or take a
screenshot);
a description of what happened and what you were doing when the problem occurred;
a description of how you tried to solve the problem;
whether you are able to reproduce the problem by repeating what you did when the problem occurred.
It may also be necessary to send the project data to SOBEK Support. The best way to do this
is to close SOBEK 3, zip the relevant project folder and send it using an e-mail. If the project
is too large to be included as an attachment in an e-mail, SOBEK Support can provide the
credentials to an ftp account where the data can be uploaded.
1.5
System requirements
The minimum hardware and software specifications for proper use of the software:
1 GHz Intel Core processor or equivalent
2 GB internal memory
10 GB free hard disk space
Screen resolution 1024x768
Microsoft .NET Framework 4.0
Recommended requirements are:
3GHz Intel Quad Core or equivalent
4 GB internal memory
20 GB free hard disk memory
Screen resolution of 1920x1080
Operating systems:
Microsoft Windows 7
Deltares
3
DR
AF
T
SOBEK 3, User Manual
4
Deltares
2 Module D-Flow 1D: Overview
D-Flow 1D is one of the models available in SOBEK 3. D-Flow 1D is the product line designed for the simulation of water flows in open channels. It combines functionality of the
former SOBEK-River Estuary and SOBEK-RIVER and is capable of modelling river systems,
estuaries, streams and other types of alluvial channel networks.
T
The software calculates accurately, fast and robust the one-dimensional water flow for shallow water in simple water systems or complex channel networks with more than thousand
reaches, cross sections and structures. D-Flow 1D solves the full Saint-Venant equations with
the help of the staggered grid numerical scheme (Stelling and Duinmeijer, 2003; Stelling and
Verwey, 2006). In order to model one-dimensional salt water intrusion in estuaries D-Flow 1D
can also solve the Saint-Venant equation and the advection-dispersion equation conjunctively
to account for advective and diffusive/dispersive transport and density driven flow.
DR
AF
D-Flow 1D allows to apply various types of boundary conditions, as well as to define lateral
inflow and outflow using time series or standard formulae. The networks can be branched
or looped. D-Flow 1D is capable of modelling complex cross-sectional profiles consisting of
multiple roughness sub-sections, e. g. left floodplain, right floodplain and main channel.
Deltares
5
DR
AF
T
SOBEK 3, User Manual
6
Deltares
3 Module D-Flow 1D: Getting started
3.1
Introduction
The workflow of setting up a D-flow 1D model usually consists of the following steps:
T
Add a D-Flow 1D model to a project
Build or import a schematization
Generate a computational grid
Define roughnesses
Set the boundary conditions
Set lateral sources and sinks (lateral stations) — if there are any
Set initial conditions
Set wind and salt values — if applicable
Adjust model wide settings
Set preferred output
Run a simulation
View and analyze simulation results
Add and combine scenarios or models - if applicable
DR
AF
These working step are explained in the following with the help of a small model without wind
data and salt water intrusion. The focus here is on workflow; an overview of the possibilities
and options of the different steps and components is provided in chapter 4.
3.2
Starting a D-flow 1D model
When SOBEK 3 is started, it opens with an empty project. To get started, import a model or
network that already exists or build a new model from scratch.
A new model is added in the Project by a right-mouse-click on <project> and chosing Add
→ New Model .... A window with all the available models from activated plugins and the
corresponding integrated models appears. Selecting Water Flow Model 1D under Hydro adds
a new flow model to the project.
The new model is now visible in the Project. Items in the Project (see also ??) are sorted
according to the usual workflow for setting up a 1-dimensional flow model as listed above.
3.3
Schematization
Selecting the Map ribbon will present all icons to add network objects to the schematization.
Always start with a channel, but we will come to that shortly. With the Map window, visualization of the network can be adjusted and map layers can be added. A wms-map layer can be
added by selecting
. After selecting “openstreetmap” the map is added to the main window.
The zoom button
, the mouse scroll-wheel and the pan zoom button
can be used to
navigate the map. Panning can also be accomplished by holding down the middle mouse
button and moving the mouse. Tip: another way to set for example OpenStreetMap as background is as follows:
Deltares
7
SOBEK 3, User Manual
right-mouse-click on Project in Project, and select Add → New Item ...
select “General” and “Map”
double click on the map
press
on top of the Map window
select “openstreetmap”
and finally right-mouse-click on the map in Project, and select “Use as default background
layer”
This way OpenStreetmap will stay as background not only while modelling the schematization
but also on presenting the Calculation grid or the Output.
T
Now, to follow this tutorial, zoom in on the city of Rotterdam as shown in Figure 3.1.
First activate an icon in the Map ribbon, then click in the Central Map (water flow 1d model(1)
window) to position the activated type of object. Start with a channel Add new branch (Freeform)
DR
AF
. Press and hold the left mouse button to place the starting point of the branch. As long as
the left mouse button is held down, the branch is drawn following the movement of the mouse
pointer. Releasing the mouse button ends the branch.
Now, to follow this tutorial, model the river section “Nieuwe Waterweg” as shown in Figure 3.1.
In this tutorial one branch is used, but more branches can be added and connected in the
same way (see also section 4.3.2.2). To stop adding branches, press Esc. Note that the order
of the mouse clicks defines the normal direction (i.e. the defined direction) of the branch,
visualized by an arrow at the end of the branch.
Figure 3.1: Map view with open street background map and a D-Flow 1D branch generated near the city of Rotterdam
Selecting
8
in the Map ribbon activates the addition mode for YZ Cross Sections. When
Deltares
Module D-Flow 1D: Getting started
Z
0
75
100
150
200
225
300
10
5
-7.5
-10
-7.5
5
10
DR
AF
Y’
T
moving the mouse over the map the orange dot shows where SOBEK places the Cross Section; with a left-click a single Cross Section is added to the branch. Press Esc to leave the
addition mode and double-click on the Cross Section in the map to open the Cross Section
Editor (Figure 3.2). In this chapter we only focus on YZ Cross Sections. The geometry of
the cross section can be specified in the table. Now, to follow this tutorial, fill in the following
values:
This will result in the Cross Section View given in Figure 3.2.
Figure 3.2: Example of a cross section
Close the Cross Section Editor and select
to add a weir. Like for the cross section, move
the mouse to a location on the branch and left-click to add a weir to the model. Leave the
addition mode by pressing Esc. A double-click on the weir opens the weir editor. Now fill in
the following values:
Deltares
9
SOBEK 3, User Manual
property
5m
200 m
DR
AF
T
Crest level
Crest width
value
Figure 3.3: Example of a weir
Close the weir-editor window and select
to add a lateral source/sink. Move the mouse
to a location on the branch and left-click to add a lateral source/sink. After pressing Esc a
double-click on the lateral node in the map or on the corresponding entry in the Project opens
the editor for lateral sources/sinks. Now set the type to Q: Constant flow and the value for the
flow to 500 m3 /s like shown in Figure 3.4.
Figure 3.4: Editor for lateral sources/sinks
The schematization now looks like Figure 3.5. Note that the extent of the Cross Section
is shown on the map. Note also that the network components are shown in the Region
window. For now, we leave the schematization as it is. For a review of all the options for
schematizations, see chapter 4.
10
Deltares
DR
AF
T
Module D-Flow 1D: Getting started
Figure 3.5: Example of the resulting schematization
3.4
Generating a computational grid
Once a schematization exists a computational grid can be generated. The computational grid
is not a part of the network, but a separate layer, which can be re-used for or linked to other
models or scenarios and redefined without influencing the network elements.
A computational grid is generated by a right-click on <computational grid> in the Project and
selecting Generate calculation grid locations. A window pops up (Figure 3.6) with a number
of options, which are described in more detail in section 4.9. For now we focus on maximum
length, which determines the distance between calculation points. Select Prefered length and
set the value to 1000 meters. After pressing OK the grid is generated and presented. For
more information on the computational grid, see section 4.9.
Deltares
11
DR
AF
T
SOBEK 3, User Manual
Figure 3.6: Computational grid editor
3.5
Boundary conditions
The boundary conditions are edited by double-clicking <Boundary Data> in the Project. In
the Central Map the boundary nodes are presented on the map and listed in a table.
12
Deltares
DR
AF
T
Module D-Flow 1D: Getting started
Figure 3.7: Boundary nodes in the Central Map
Right-mouse-clicking on one of the nodes in the table and selecting Open View ... opens an
editor. The following types of boundary conditions can be selected:
None
H(t): Waterlevel time series
Q(t): discharge time series
Q(h): discharge waterevel table
Q : constant discharge
H : constant waterlevel
Now, to follow this tutorial, select a constant flow of 800 m3 /s at the start of the branch (most
upstream point), and a constant waterlevel of 1 m at the end of the branch (most downstream
point), as shown in Figure 3.8.
Figure 3.8: Constant water level boundary condition
3.6
Roughness
Branch roughness can be defined for different parts of the cross sections, defined as roughness-sections. Open the Cross Section Editor to view the definition of the sections, in the
Deltares
13
SOBEK 3, User Manual
table underneath the graphical representation. The roughness-section is visualized by the
block under the cross section in the graphical representation. For the simple model discussed
in this chapter, keep a single roughness-section.
DR
AF
T
Notice that the model wide roughness-type and -value can be edited in the Properties window
after selecting Main under <water flow 1d (1)/input/Roughness> (Figure 4.43). Press the "+"
in front of <Roughness> to unfold. For now, in this simple model, the default roughness is
not changed and no detailed roughness value is defined for the branch. More information on
setting roughness is found in section 4.6.
Figure 3.9: Editing the roughness
3.7
Initial conditions
There are two basic initial conditions:
initial waterlevel or depth; and
initial water flow (discharge)
Both can be specified. The user can choose between initial waterlevel or depth by selecting
the <Water flow 1d (1)> model in the Project window and the Initial conditions section in the
Properties window.
14
Deltares
DR
AF
T
Module D-Flow 1D: Getting started
Figure 3.10: Output options in the Properties Window
Now, for this tutorial, change the definition from Depth to Waterlevel, then set its Default value
to “1”. Note that in the Project <initial water depth> has now changed to <initial waterlevel>.
Leave the initial water flow as it is.
3.8
Model parameter settings
Some parameters need to be set before a model run. By selecting Project <water flow 1d
(1)>, the simulation settings for the model appear in the Properties window. There are several parameters, which can be edited, but the most important are StartTime, StopTime and
TimeStep. The parameters StartTime and StopTime define the simulation period. The parameter TimeStep defines the maximum time step with which the simulation is performed. Whenever and wherever in the schematization the numerical scheme requires a smaller timestep
to ensure computational stability, the program will reduce the timestep as necessary. Please
note that the automated reduction of timestep is only done to prevent model crashes. Based
on the modelled hydrodynamic phenomena, users should select appropriate space-steps as
well as an appropriate timestep to ensure that the hydrodynamic phenomena involved are
computed with sufficient accuracy.
Now, to follow this tutorial, set the simulation period to 3 d by adjusting <Start time> and
<Stop time>. Set the <Time step> to 1 h.
Deltares
15
SOBEK 3, User Manual
Set output
Left-click on Project “water flow 1d (1)/output”. The Properties Window now shows all possible output options, see Figure 3.11.
Choose the following output parameters and set the output value on “Current”:
Grid points: Water level
Reach segments: Discharge
Reach segments: Velocity
Simulation info: Number of iterations
Structures: Crest level
T
Set the rest of the parameters to “None”. Set both output timesteps to 1 h.
DR
AF
3.9
16
Deltares
DR
AF
T
Module D-Flow 1D: Getting started
Figure 3.11: Output options in the Properties Window
3.10
Validation
As a final step in the modelling process, the user can activate the validation tool, by rightmouse-click in the Project on <water flow model 1D (1)> and selecting Validate.... The
validation tool checks all that is required for a model run. In other words: a validated model
will run!
Deltares
17
SOBEK 3, User Manual
3.11
Running a simulation
The simulation can now be started by a right-mouse-click on Project <water flow 1d (1)>
and selecting Run model. A window pops up in which the progress of the simulation is shown.
The window disappears when the simulation is finished.
In the Project window the results are added to the model under <Output>. Here the run
report can also be found. This is a log with all the messages during the simulation from both
Delta Shell and D-Flow 1D.
Viewing simulation results
T
The simulation results can now be examined and analyzed in several ways, described in detail
in Chapter chapter 5.
Double-clicking on Project <output/water level> will present the water level results on the
map together with the network as coloured symbols on the calculation points. The initial
values of the water level will be presented first. Now, activate the Time Navigator. You can
move through the results as a function of time by moving the slider (Figure 3.12). On the
left side of the map view, the results in table are visible for the specified time in the Time
Navigator window.
DR
AF
3.12
Figure 3.12: Map results of water level
Press and hold the Ctrl key, then left-click on three locations in the map view of the water
level results. The three locations are all selected. Make sure you choose a value upstream of
the weir node, downstream of the weir node and downstream of the lateral source/sink node.
Now left-click on the Query Time Series icon
in the Map ribbon to get the time series of
the results. A window pops up in which you can choose one or more parameters. Select water
level and press OK. Note that it is possible to choose a parameter different than water level
even though the locations were selected in a water level map view. A new tab now opens with
18
Deltares
Module D-Flow 1D: Getting started
a graphical view (the Function view) of the requested time series on the right and a table
with the depicted results on the left (Figure 3.13).
DR
AF
T
Returning to the map view and selecting new locations and an output parameter adds a new
line to the Function view after clicking the Get time series icon. Change the curves in the
Function view in the Chart window (Figure 3.14).
Figure 3.13: Results of water level for three locations along the branch in Function view
Deltares
19
DR
AF
T
SOBEK 3, User Manual
Figure 3.14: Chart and the corresponding Properties window
20
Deltares
4 Module D-Flow 1D: All about the modeling process
4.1
Introduction
4.2
4.2.1
Import
Network (schematization)
Boundary conditions
Initial Conditions
Roughness
Wind data
Salinity
Computational Grid
Parameter settings
Validation
Output
DR
AF
T
This chapter describes the functionality of the D-Flow 1D plug-in:
Import
Import modeldata on <Project> level
On the project level, data from other models or full Delta Shell projects can be imported by
a right-mouse-click in the Project window on <Project> and choosing Import. . . . Figure 4.1
shows the resulting pop-up screen. Different data types can be imported:
a Project (SOBEK 3)
a network from a geographical information system (GIS)
a NetCDF regular two-dimensional grid
a Raster File
a Time series (CSV)
a Time-dependent grid
a SOBEK model or network (works for SOBEK-RE and SOBEK 2)
Deltares
21
DR
AF
T
SOBEK 3, User Manual
Figure 4.1: Data Import window
In case of importing a SOBEK model the user can choose between different options. Note
that the data will be stored at the Project level. This implies that for example the imported
network is not connected to any existing model in the Delta Shell project. At a later stage the
network can be linked by dragging the network onto the flow model 1d in the Project window.
4.2.2
Import a network from another model on <network> level
In case of an existing D-Flow 1D model, the network, the network objects and cross sections
(with profile data) can be imported to complete or update the model on the <network> level.
Right-click on <Project/flow model 1d / input / network> on an existing D-Flow 1D model in
the Project window. Figure 4.2 shows the resulting pop-up screen. The following options are
available:
SOBEK data
Model features from GIS
data in text files (<csv>) for three types of cross sections
22
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.2: Data import window for network (features)
By selecting the appropriate import a wizard window pops up. After completing the wizard,
the data is imported and added to the Project window. Data already existing in the D-Flow
1D model is overwritten with the imported data.
4.2.3
4.2.3.1
Import a network from GIS
The GIS import wizard
To import a network with its objects a GIS import wizard is available. The wizard can be
addressed on a <network> level (section 4.2.2) or on a <project> level (section 4.2.1).
The wizard imports the network itself or network features from a shapefile <shp> or from
a personal geodatabase <mdb>. Always start with the Channels. Figure 4.3 shows the GIS
import wizard.
Deltares
23
DR
AF
T
SOBEK 3, User Manual
Figure 4.3: The GIS import wizard
A complete network consists of a combination of different network features from several
shapefiles or tables in a personal geodatabase. Here, the description is limited to the import
from shapefiles. In section 4.2.3.2 the specifics of importing from a personal geodatabase
is described. Several network objects can be imported simultaneously by selecting Features
with the <shp>-file and adding it (
) to the (import)list (Figure 4.3). When all
the required features are set, click Next. Another window appears with the mapping table
(Figure 4.4).
24
Deltares
T
Module D-Flow 1D: All about the modeling process
DR
AF
Figure 4.4: Example of the mapping table
Here, columns in the shapefiles can be related to D-Flow 1D network objects. After defining
the mapping and confirmation with Next the Import properties window appears where the
user can set the snapping precision (Figure 4.5). Network objects like weirs or cross sections
in the <shp>-file will be snapped to the nearest branch during import, unless they are farther
from any branch than the snapping precision specified in the Import properties window. In
the latter case, the network objects are discarded during the import.
Figure 4.5: Import properties window for snapping precision and saving of mapping files
Another important button in this screen is
. A mouse-click on this button
saves the entire mapping of the different shapefiles. Instead of having to walk through all
above-mentioned steps to import a network, the next import from a similar set of shapefiles
(or personal geodatabase) can be handled by a mouse-click on
Figure 4.3.
Deltares
in
25
SOBEK 3, User Manual
4.2.3.2
Import from personal geodatabase
To import network objects from a personal geodatabase, the user must activate the GIS import
wizard, described in the previous paragraph. Here, the specifics on importing from a personal
geodatabases are described.
In case the user selects a personal geodatabase in the first screen of the GIS import wizard,
the user must also set the correct feature class in Table before adding the feature to the
import-list.
DR
AF
T
Some features have additional tables which have to be taken into account. For example, cross
sections often have separate tables for the profile data. Relating tables can be added by a
mouse-click on
in the GIS import window. The relations between the base table and a
related table are set in another window, an example is given in Figure 4.6. Similarly to joining
of tables in ESRI ArcGIS, the related tables and the matching ID-column are set. It is also
possible to filter specific columns or values.
Figure 4.6: Setting of related tables
Mark, that adding features to the import-list, mapping, snapping and import all work the same
as for shapefiles.
4.2.3.3
Import of culvert (profile) data
In case of an existing D-Flow 1D model network, GIS data for culverts can be imported. The
˘
˘ Z´ is applied, see http://www.aquo.nl/aquo/lm_aquo/element/
âAŸAQUO
standaardâA
KDUVORM.htm. The following shapes are recognized (dutch):
26
Round - default
Rectangle
˘
˘ Z)
´
Egg (âAŸeivormigâ
A
˘
˘
´
Cunette (âAŸmuilâAZ)
Ellips
˘
˘ Z)
´
Arch (âAŸheulâ
A
–: default
Deltares
Module D-Flow 1D: All about the modeling process
˘
˘ Z´
The latter type is converted to a round culvert. Of course, to achieve this the âAŸshapeâ
A
˘
˘
´
˘
˘
´
˘
˘
´
must be mapped. In addition, the âAŸheightâAZ, âAŸwidthâAZ and âAŸdiameterâAZ of a
culvert can be mapped as well. These will be recgnised and imported into the D-Flow 1D
model network.
Import cross section profiles from <csv>
T
Cross sections (location and profile) can be imported from <csv>-files. This can be done
either by a right-mouse-click in the Project window on <Project / water flow model 1D / input
/ network> and selecting Import ... or by a right-mouse-click in the Central Map and selecting
Import cross section(s) from .csv. After selecting the <csv>-file, the following window popsup:
DR
AF
4.2.4
Figure 4.7: Example importing YZ Cross Section from <csv>-file
The screenshot above shows the columns SOBEK 3 requires. An example file can be obtained
by exporting some cross sections.
By default, cross sections with the same Name will be replaced. By de-selecting Import
chainages the location of the original cross section can be left unchanged. In that case, the
column can be left empty. By default, the option Create cross section if Name was not found
in the network is activated.
Note: that the import from <csv>-file described here can be used to replace the (profile)data
of present cross sections. This way, the import of cross sections is a two-step process:
first import the cross sections (location) by the GIS import wizard (see section 4.2.3) - a
default profile will be added;
then import the cross section profiles from <csv> described here. De-select Import
Deltares
27
SOBEK 3, User Manual
chainages.
Import time series from <csv>
T
A timeseries of waterlevels or discharges can be imported by a right-mouse-click in the
Project window on <Project> and selecting Import... and Time series (CSV). A wizard
opens in which a <csv>-file can be selected and the delimiters between the columns can
be set (Figure 4.8).
DR
AF
4.2.5
Figure 4.8: Selecting delimiters for a csv file
The columns with the date-time and the data are specified as shown in Figure 4.9.
28
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.9: Selecting the columns of the <csv>-file
The timeseries is added to the project and can be used as a boundary condition or lateral source. Link the timeseries by selecting and dragging it in the Project window onto a
<Boundary Data / Node...> or onto <Lateral Data / LateralSource...>.
Deltares
29
DR
AF
T
SOBEK 3, User Manual
Figure 4.10: Linking a Timeseries
4.3
4.3.1
Network
Setting up a network from scratch
To build a model schematization from scratch, add a new model to the project (right-mouseclick on <project> in the Project window, select Add → Water Flow model 1D)
By double-clicking on <network> in the model input in the Project window, the network will
be presented in a map (central workspace). In the Map ribbon (??) there are several buttons
to add network objects like
30
branche
node
cross section
structure: weir, pump, culvert, bridge
extra resistance
Deltares
Module D-Flow 1D: All about the modeling process
retention
lateral source/sink
observation point
4.3.2
4.3.2.1
node
cross section
weir
pump
culvert and syphon
bridge
extra resistance
retention
lateral source/sink
observation point
DR
AF
T
are activated. Pressing the Esc key ends the editing mode, the selection tool is activated.
Double-clicking on a network element either in the Central Map or in the Region window
opens the corresponding editor in a new tab. A network consists of point elements and line
elements. Branches are the only type of line elements, but there are multiple point elements:
Nodes and branches
Nodes
Nodes are the basis of any network. They define the limits of branches and the network itself.
If a node forms the boundary of the network, often a boundary conditions is set at the node.
Moving a node
therefore changes the length and geometry of branches. The location
of a node is defined by x-y -coordinates which can be adjusted in the Properties window of
the node.
Deltares
31
DR
AF
T
SOBEK 3, User Manual
Figure 4.11: Example of boundary nodes
A node can belong to a single branch - where it will limit the network - or to more branches.
Figure 4.11 shows an example of two connected branches. The node connecting the two
branches is solid green. The connection node works as a boundary between branches. The
characteristics of the branches are not interpolated across the connection node. Instead, the
waterlevel and discharge are transfered from one branch to the next. When selecting the
branch, the curvepoints are shown as green squares. The branch direction is shown by a blue
arrow. Most users will use the direction of the flow. The branch direction can be reversed by
right-clicking the branch in the map and selecting Reverse direction.
The table below the Central Map, see ??) contains a tab with all nodes in the network. Nodes
can be selected in the map and in the Attribute Table. Nodes are added automatically when
a new branch is drawn, but can also be added or removed by a right-mouse-click in the Central
Map and selecting Insert node or Remove node. Boundary nodes can not be removed.
4.3.2.2
Branches
A branch is a line object between two nodes. With a branch, the course of a river, channel or
stream is schematised. A branch always has a geometry and hydraulic characteristics:
A start and end location (nodes), which determine the boundaries of the branch and the
length
Curvepoints (which set the curvature-geometry)
Dimensions (cross sections)
A resistance (hydraulic roughness)
In addition, there can be additional branch-features, such as structures or lateral sources.
For adding branches the Map ribbon (??) of D-Flow 1D provides several tools. Create new
branches
32
Deltares
Module D-Flow 1D: All about the modeling process
by point and click
or
automatic curve points
.
In this editing mode an additional branch can be connected by re-using a node of the existing
branch. Two existing branches can be connected by drawing a new branch using the nodes
of the existing branches. These nodes will change to solid green.
Reposition an existing branch
by adding curve points with
moving a single curve point with
.
T
or moving the branch as a whole from a selected curve point with
4.3.2.3
DR
AF
These move features can also be used for other network elements. To reverse the branch
direction right-click the branch in the map with the mouse and select Reverse direction.
Interpolation across nodes
By default, the characteristics (cross section, bed level and roughness) of a branch are not
interpolated across a connection node. Instead, the waterlevel and discharge are transfered
from one branch to the next. In case of a single river or channel this might lead to unrealistic
and undesired flow.
Figure 4.12: Two branches with different Order number: No interpolation across the connection node
To avoid this the Order number of branches is introduced. Cross sections, bed level and
roughnesses are interpolated across a connection node when the branches have the same
Order number. By specifying the same Order number for branches of the main river, a tributary can be distinguished by a different Order number. The characteristics of the main river
will be interpolated, resulting in a smooth flow. This can be achieved as follows:
Press the Esc key
Select the first branch by a left-click on the map
Hold the Ctrl key while selecting the next one; or
Deltares
33
SOBEK 3, User Manual
Hold the Shift key while selecting another branch: the shortest route will be selected
(the selection will be high-lighted)
While holding the Ctrl key, a left-click will de-select a branch
DR
AF
T
Now the user can modify the Order number of all selected branches in the Properties window.
Figure 4.13: Two branches with same Order numbers: Bed level is interpolated across
the connection node
In case the user starts to model from scratch, the Order number of branches is set to "-1" (no
interpolation). It is advised to change the Order number, as the application will then apply the
Order number for new branches, according to the following rules:
a new (continuous) branch gets the same Order number
a new branch at a junction gets a higher Order number
after splitting a branch the new branch gets the same Order number
All branches of an imported SOBEK 2 model will have the default Order number of "-1". For
compatability the Order numbers of branches are adjusted in line with the concept of Linkage
node. This concept has been discarded in SOBEK 3.
4.3.3
4.3.3.1
Weir
Introduction
A weir is a point object on a branch that limits the flow in that branch by adding a physical
blockage with certain dimensions, representing a hydraulic structure in the real world. Weir
objects can model three types of flow:
free flow
submerged flow
no flow
A weir can be simple or more complicated, which results in the following types (where the
term in brackets is the corresponding name in SOBEK 2):
34
Deltares
Module D-Flow 1D: All about the modeling process
Simple weir (Weir)
Gated weir (Orifice)
Weir with piers (Advanced weir)
Weir with detailed description of crest (River weir)
Free form weir (Universal structure)
General structure
Each type of weir has a specific shape and parameters to be set. For a detailed description of the underlying mathematical model we refer intermediately to the technical reference
(Deltares, 2012).
T
A weir can be added to the network by clicking on
in the Map ribbon (??). Then click on
the preferred location in the network to position the weir. The weir is snapped to the nearest
location on a branch. A second way to add a weir to the network is by right-mouse-click in
the Region window, on the branch, and select select Add Weir. The weir is added at zero
chainage. This can be adjusted in de Properties window.
DR
AF
Double-clicking the weir object in the Central Map or in the Region window opens the weir
editor in a new tab. The editor window has the following elements:
a graphical representation of the structure in side view and in cross section view
a tab with the structure-ID, which can be used in composite structures to switch easily
between structures in the composite
structure settings.
Some of the structure properties can also be edited in the Properties window or the Attribute
Table.
4.3.3.2
Simple weir
The editor for a simple broad-crested weir is shown in Figure 4.14. For a simple weir the
following parameters can be adjusted in the editing window:
Crest level: the height of the weir crest in meters
Crest width: the width of the weir in meters
Allowed flow direction, Positive
Allowed flow direction, Negative
Discharge coefficient Ce (dimensionless): the default value is 0.8.
Lateral contraction coefficient Cw (dimensionless): the default is 1.0
Deltares
35
DR
AF
T
SOBEK 3, User Manual
Figure 4.14: Simple weir editor
4.3.3.3
Gated weir
The editor for a gated weir is shown in Figure 4.15. Editable parameters for a gated weir are:
Crest level: the height of the weir in meters
Crest width: the width of the weir in meters
Lower edge level: the lower edge of the gate in meters
Gate opening: distance between weir crest and gate lower edge in meters
Allowed flow direction, Positive
Allowed flow direction, Negative
Max: a maximum discharge [m3 /s] for both the positive and negative allowed flow directions
Discharge coefficient Ce : default is 0.8
Lateral contraction coefficient Cw : default is 1.0
The lower edge gate level is automatically set when the crest level and/or the opening height
are adjusted. Similarly, when the lower edge gate level is adjusted, the gate opening is automatically adjusted as well.
36
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.15: Gated weir editor
4.3.3.4
Weir with piers
The editor for a weir with piers is shown in Figure 4.16. Editable parameters for a weir with
piers are:
Crest level: the height of the weir in meters
Crest width: the width of the weir in meters
Number of piers
Upstream face P : height of the weir relative to the bed level at the upstream side in meters.
The default value is 10 m
Design head of weir flow H0 : the head for which the structure was designed. The default
value is 3 m
Pier contraction coefficient Kp : coefficient representing the net sill-width reduction due to
the presence of piers. The value depends on the shape of the piers, the default value is
0.01
Abutment coefficient Ka : coefficient representing the net total flow width reduction due to
the presence of abutments. The value depends on the shape of the abutments, default
value is 0.01
Deltares
37
DR
AF
T
SOBEK 3, User Manual
Figure 4.16: Weir with piers editor
4.3.3.5
Weir with detailed description of crest
For a “weir with detailed description of crest” the crest shape can be set in addition to the crest
level and the crest width. In a drop-down menu the user can choose between:
Broad
Sharp
Round
Triangular.
Figure 4.17 shows the editor. In the side-view the structure shape changes with the type
of crest. Furthermore, the following energy loss properties must be specified for each flow
direction:
Correction
Submerge
Reduction table: reduction coefficient as a function of the head
Default values are provided.
38
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.17: Weir with detailed description of crest editor, the side-view shows the shape
of the crest
4.3.3.6
Free form weir
Free form weirs can be defined by a Y-Z profile. The weir consists of rectangular sections
having a horizontal bed and triangular sections having a sloping bed. It is assumed that the
total discharge over a free from weir is the sum of the discharge over each section, where
rectangular weir sections are considered as a simple weir and a triangular weir sections are
considered as (the half of) a broad-crested weir with truncated triangular control section.
Figure 4.18 shows a typical free from weir, suitable for fish to pass the weir - even at low
discharge. The following parameters can be adjusted in the editing window:
Y’, Z table (minumum of 2 values)
Allowed flow direction, Positive
Allowed flow direction, Negative
Discharge coefficient Ce (dimensionless): the default value is 0.8.
Deltares
39
DR
AF
T
SOBEK 3, User Manual
Figure 4.18: Free form weir editor
4.3.3.7
General structure
This type of weir is a special kind of gated weir with additional information on the geometry
and the possibility of drowned gate flow and drowned weir flow. For more information see
Technical Reference Manual. Figure 4.19 shows the editor. Editable parameters are
Lower edge level: gate lower level in m AD
Gate opening: height in m
Level and width: table with levels in m AD and widths in meters for upstream location 1
and 2, and downstream locations 1 and 2, see for a detailed explanation the technical
reference manual.
Coefficient free gate flow: coefficient representing the contraction for free gate flow
Coefficient drowned gate flow: coefficient representing the contraction for drowned gate
flow
Coefficient free weir flow: coefficient representing the contraction for free weir flow
Coefficient drowned weir flow: coefficient representing the contraction for drowned weir
flow
Contraction coefficient
40
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.19: General structure editor
4.3.4
Pump
To add a pump object click on
in the Map ribbon (??). Then click on the preferred location
in the network to position the pump. The pump is snapped to the nearest location on a
branch. A second way to select add a pump to the network is by right-mouse-click in the
Region window, on the branch, and select select Add Pump. The pump is added at zero
chainage. This can be adjusted in de Properties window.
By double-clicking on the pump in the Central Map or in the Region window, the pump editor
is opened in a new tab. The pump editor is shown in Figure 4.20. For a pump the editable
parameters are
Pump capacity in m3 /s
Pump direction (positive or negative)
Switch-on and -off levels for both the suction side and the delivery side. The switch-on
levels are depicted by a black line in the cross section view, the switch-off levels by a red
line
A reduction table can be specified optionally
Deltares
41
T
SOBEK 3, User Manual
DR
AF
Figure 4.20: Pump editor
In D-Flow 1D a pump can only have one capacity and one set of switch on/off levels. A pump
with multiple capacities and multiple switch on/off levels is modelled as a composite structure
(see section 4.3.6) consisting of several pumps.
4.3.5
Culvert, Syphon and Inverted Syphon
In order to model pipe-shaped structures that connects two open channels, for example a pipe
underneath a road connecting two waterways, D-Flow 1D provides three different structure
features:
Culvert
Syphon
Inverted Syphon
Culvert, Syphon and Inverted Syphon can be equipped with gates. The discharge through a
culvert is affected by the upstream and downstream invert levels, its shape, size and length
and the material.
A Culvert can be added by clicking
in the Map ribbon (??). Then click on the preferred
location in the network to position the Culvert. The Culvert is snapped to the nearest location
on a branch. A second way to add a culvert to the network is by right-mouse-click in the
Region window, on the branch, and select select Add Culvert. The culvert is added at zero
chainage. This can be adjusted in de Properties window.
By double-clicking on the culvert in the Central Map or in the Region window, the Culvert
Editor is opened in a new tab.
42
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.21: Culvert editor
Parameters that can be specified are:
Length: length of the culvert in m
Groundlayer: roughness type and value. For roughness type the options are
Chézy (C )
Manning (nm )
Strickler (kn )
Strickler (ks )
White and Colebrook
Geometry type
Tabulated
Round
Egg
Rectangle
Ellipse
Arch
Cunette
SteelCunette
In the Culvert editor it is also possible to check the
box. The Culvert is then treated
as a syphon with an On/Off level. The On level and the Off level are displayed in the side-view
of the editing window. The user also has to specify a Bend loss coefficient unequal to 100. In
addition, the syphon may be inverted. By unchecking the
Deltares
box but leaving the Bend
43
SOBEK 3, User Manual
loss coefficient unequal to 100, the culvert is treated as an Inverted Syphon. And finally, a
gate can be added by checking the
edge level.
Composite structure
T
A composite structure is a combination of multiple structures of the same type or different
types. D-Flow 1D distributes the water to the structures according to the mathematical models
of the structure objects. An example of a composite structure (also: compound structure)
is given in Figure 4.22. In the Region the structure objects forming together a D-Flow 1D
Composite Structure are summarized under StructureFeature (Figure 4.23). In the Properties
window of a Composite Structure the number of structure objects is displayed. To create a
composite structure, add multiple structure objects to the same location.
DR
AF
4.3.6
, and specifying Initial gate opening and Lower
Figure 4.22: Example of a Composite Structure in the Central Map. Multiple structure
objects (here: two Weirs, a Pump and a Culvert) are arranged horizontally,
the bar below the structure icons indicates that the structures are combined
to a Composite Structure. The Attribute Table lists the sub-structures of
Weir type
Figure 4.23: Region window with a Composite Structure consisting of two weirs, a pump
and a culvert
44
Deltares
Module D-Flow 1D: All about the modeling process
Bridge
A bridge forms a resistance for water flow that depends on the cross section under the bridge
and the shape of the pillars. There are three types of bridges (where the term in brackets
corresponds to SOBEK 2 terminlogy):
Rectangle (fixed-bed and soil-bed bridge)
Tabulated (abutment bridge)
Pillar (pillar bridge)
DR
AF
T
Add a Bridge to the model network by clicking the Add Bridge tool
in the Map ribbon (??).
Then click on the preferred location in the network to position the Bridge. The Bridge object is
snapped to the nearest location on a branch. A second way to add a bridge to the network is
by right-mouse-click in the Region window, on the branch, and select select Add Bridge. The
bridge is added at zero chainage. This can be adjusted in de Properties window.
Figure 4.24: Bridge editor
By double-clicking on the Bridge in the Central Map or in the Region window, the bridge editor
is opened in a new tab. The bridge editor is shown in Figure 4.24. For a bridge the editable
parameters are:
Geometry of the (cross sectional) flow-area, choose between
4.3.7
Rectangle: specify the cross section geometry in a table
Tabulated: specify the cross section geometry in a table
Deltares
45
SOBEK 3, User Manual
Pillar: fill in the fields for the width between the pillars and the shape factor
Length: the length of the bridge along the course of the river in [m], displayed in the
side-view
Roughness Type: choose between
4.3.8
the corresponding roughness value (the unit depends on the roughness type)
a Ground layer roughness option
Allowed flow direction (positive, negative or both)
Inlet loss and Outlet loss
Extra Resistance
T
Chezy
Manning
Strickler (kn and ks )
White-Coolebrook
An Extra resistance object can be used to model sill beams or other obstacles in the channel
DR
AF
not further specified or to to adjust the water distribution in a bifurcation. By clicking
in
the Map ribbon (??), the user can add an Extra resistance object. A double-click on the
Extra resistance object in the Central Map or in the Region window opens the editor with the
following editable parameters:
Choice of two formulas to compute the extra resistance
A table that defines the extra resistance parameters in dependence of the waterlevel
4.3.9
Lateral Source
A lateral source (sink) is a volume of water entering (leaving) the model at a location on
a branch within a certain period of time. As a sink can be interpreted as a source with a
negative sign, the corresponding object in D-Flow 1D has been named “Lateral Source”.
To add a Lateral Source click
in the Map ribbon. Then click on the preferred location in
the network to position the Lateral Source. The object is snapped to the nearest location on
a branch. A second way to add a Lateral Source to the network is by right-mouse-click in the
Region window, on the branch, and select select Add Lateral. The Lateral Source object is
added at zero chainage. This can be adjusted in de Properties window. Mark the difference
between Lateral source on selection in the Region window, and Lateral source boundary data
on selection in the Project window.
Now that the Lateral Source is positioned on the network, the volume of water can be defined
either as constant or as a function of time or waterlevel.
46
Deltares
Module D-Flow 1D: All about the modeling process
T
Figure 4.25: Editor for lateral source data
DR
AF
A time series can be generated by a right-mouse-click in the Project window on <Lateral
Data / LateralSource..> and selecting Generate data in series.... Figure 4.26 shows the popup screen, where Start, End and Interval can be set. By clicking on Generate data a table is
generated with a constant discharge. The user can change the discharge values by opening
the editor for lateral source data (Figure 4.25). This editor is evoked with a double-click in
the Project window on <Lateral Data / Lateral Source...>. A positive value for discharge
represents water flowing into the system, a negative value means water flowing out of the
system. It is good modeling practice to limit the lateral inflow or outflow to 10 % of the channel
flow.
Figure 4.26: Generate data series
It is also possible to use a Timeseries which is available in the <Project>. Chapter section 4.2.5 describes how a Timeseries can be imported and linked to a Lateral source.
Deltares
47
SOBEK 3, User Manual
4.3.10
Retention area
By clicking
in the Map ribbon (??) the user can add a Retention area object. The Retention
area parameters can be edited in the corresponding Properties window.
4.3.11
Observation point
Clicking the Add Observation Point
in the Map ribbon (??) allows to add an Observation
Point to the network. At Observation Points the simulated
discharge
velocity
depth
waterlevel
T
4.3.12
4.3.12.1
DR
AF
can be visualized at a smaller time step than for the calculation points (see chapter 5). Observation Points are often used as input locations for D-RTC flow charts or they represent a
gauge in the real-world river system and so become a location of interest.
Cross Section
Adding Cross Sections to the network
A Cross Section is added in two steps. First, select the type of cross section by activating one
of the following tools from the Map ribbon (??):
Add Cross Section YZ
for a cross section with a default definition. This default definition must be specified
previously in the Region window as described in section 4.3.12.6.
Add Cross Section ZW
Add Cross Section XYZ
Add Cross Section
with a rectangle, arch, (steel)cunette, ellipsis or trapezium profile
Second, add the Cross Section by clicking on the preferred network location. The Cross
Section object snaps to the nearest location on a branch. To leave the network editing mode,
press Esc.
It is also possible to add a cross section to the network by right-mouse-click in the Region
window, on the branch, and selecting Add Cross Section YZ ... or Add Cross Section ZW ....
In the pop-up window the chainage and Z Level shift can be specified.
The cross sections can now be edited by double-clicking on the Cross Section object in the
Central Map or in the Region window, or by selecting the Cross Section in the Attribute
Table below the Central Map, right-mouse-clicking and selecting Edit. This is described in
more detail in the following parapraphs.
48
Deltares
Module D-Flow 1D: All about the modeling process
T
Cross Section YZ
DR
AF
4.3.12.2
Figure 4.27: Cross Section editor for YZ Cross Sections with table (left) and graphical represenation of the cross section geometry. The vertical line indicates where
the cross section crosses the branch line. The two highlighted points in the
diagram correspond with the selected row in the table. Below the graphical
representation is the table the roughness information (see section 4.6)
Figure 4.27 shows the editing window for Cross Sections. On the left there is a table with
the yz -coordinates of the cross section. On the right a graphical representation of this table
content is given. The cross section geometry can be modified in the table or in the diagram.
While navigating in the table the points corresponding with the active row are highlighted in
the diagram. Extra storage volume can be created by adding a positive value in the column
∆z Storage of the yz -table or dragging the points in the diagram - upwards only. The storage
volume is visualized in the graph as a shaded area. This part of the cross section is not
considered as cross sectional flow area. In the diagram the cursor switches automatically
from add mode to drag mode. In case of zero storage the Total and Flow profile are equal
(double line). Hold the Alt key and drag to modify both profiles. Move the mouse while holding
the left mouse button pressed to the right and down to zoom in, and to the left and down to
zoom out. Move the dotted vertical line to shift the cross section with respect to the branch
line (thalweg). Note that also the roughness sections are defined in the Cross Section editor,
for a full description see section 4.6.
Deltares
49
SOBEK 3, User Manual
T
Cross Section XYZ
DR
AF
4.3.12.3
Figure 4.28: Editing window for an XYZ Cross Section
Cross Sections of XYZ-type are similar to YZ Cross Sections, but they are usually drawn
directly on the map, so the cross section points are not necessarily arranged on a line orthogonal to the branch line. Figure 4.28 shows the editing window for an XYZ Cross Section.
The editing window for XYZ Cross Sections is similar to the one for YZ Cross Sections (section 4.3.12.2), but the table shows y 0 values in the first column. These are the projected values
along a straight line as shown in Figure 4.29. As SOBEK is a 1D model, the geometry has to
be projected to a single location on the branch. This projection is length-conserving; the total
length of the cross section is maintained. The first location has offset 0, the end location has
offset L.
Figure 4.29: Projection of a xyz-cross- section
50
Deltares
Module D-Flow 1D: All about the modeling process
Use the Move Feature tool
the table can not be edited.
Cross Section ZW
T
ZW Cross Sections are mainly used in the modeling of rivers. They correspond to the Tabulated River Cross Sections in SOBEK 2. They are usually calculated by external software (for
example BASELINE/WAQ2Prof) and imported into a flow schematization. Figure 4.30 shows
the editor for ZW Cross Sections. Instead of a location-level relation a ZW Cross Section has
a relation between the channel width and the waterlevel. In addition, there is a difference
between the flow width (the part of the channel that takes part in the actual flow) and the total
width (the flow width with additional storage). As a consequence, ZW Cross Sections are
always symmetrical.
ZW Cross Sections can incorporate a summer dike with additional flow and storage area
(Figure 4.30). The part of the floodplain behind the dike does not play a role in the computation
until the waterlevel exceeds the crest level of the summer dike. When a summer dike floods
the extra area is added to the cross section. To prevent the flow area from taking part in the
flow process too easily, D-Flow 1D uses a transition height (see section 4.10) above the crest
level to ’scale’ the flow into the floodplain. When the waterlevel falls below crest level, the
extra area is gradually removed again from the cross section, modeling the water behind the
summerdike to flow back slowly into the river until the flood plain is dry again.
DR
AF
4.3.12.4
to adjust the points in the horizontal plane. The y 0 values in
Figure 4.30: Cross section editor for ZW Cross Sections
Deltares
51
SOBEK 3, User Manual
4.3.12.5
Cross Section
Cross Sections allow to specify simple geometries like:
Rectangle
Arch
Cunette
SteelCunette
Ellips
Trapezium
DR
AF
T
Figure 4.31: Cross section editor for Trapezium
In the editing window for Cross Sections these geometries can be defined. Figure 4.31 shows
the Cross Section editor with a trapezium cross sectional profile as example. It is not possible
to model storage volume with these types of cross sections.
4.3.12.6
Working with Shared Cross Section definitions
The geometry (profile) and other parameters can be shared with different Cross Section objects in a network. Modifying a Sshared Cross Section will change the definition, and therefore
change the cross section data of all Cross Sections that refer to the Shared Cross Section
Definition.
Note: that the level-shift is not shared. It is specified for each Cross Section object individually
in the Cross Section editor (Figure 4.32).
To make an existing cross section sharable, open the Cross Section editor, choose use local
definition and press Share this definition.
52
Deltares
Module D-Flow 1D: All about the modeling process
Figure 4.32: Switch between Local Cross Section definition and Shared Cross Section
definition in the Cross Section editing window
Now, the cross section can be used at different locations in the network. Shared Cross Section
Definitions are listed in the Region window.
Note: Several options are available, by a right-mouse-click in the Region window, under
<Shared Cross Section Definitions>, on a cross section:
Rename
Delete
Show usage. . . lists locations where this Shared Cross Section Definition is used.
Set as default. Now, the user can add a default cross section by selecting the Add Cross
Section from Shared Default Definition
T
in the Map ribbon (??).
DR
AF
Quick fix: Place on empty branches will place the Shared Cross Section Definition on
branches which do not yet have any cross section
4.3.12.7
Import and export cross sections from/to <csv>-file
Cross sections (location and profile) can be imported from <csv>-files. This can be done
either by a right-mouse-click in the Project window on <Project / water flow model 1D / input
/ network> and selecting Import ... or by a right-mouse-click in the Central Map and selecting
Import cross section(s) from .csv. After selecting the /extcsv-file, the following window popsup:
Deltares
53
DR
AF
T
SOBEK 3, User Manual
Figure 4.33: Example importing YZ Cross Section from <csv>-file
The picture above shows the columns SOBEK 3 requires. An example file can be obtained by
exporting some cross sections.
By default, cross sections with the same Name will be replaced. By de-selecting Import
chainages the location of the original cross section can be left unchanged. In that case, the
column can be left empty. By default, the option Create cross section if Name was not found
in the network is activated.
The export of cross sections works the same way. The cross sections can be exported,
modified outside SOBEK and then be imported again. If a Cross Section with the same name
or id already exists, this Cross Section is updated with the values from the imported file. If a
Cross Section with the same name or ID is not present in the network, it is added as new.
4.3.12.8
Inspect multiple cross sections in one view
It is possible to inspect multiple cross sections in one view, as follows:
unfold/expand the cross sections which you want to inspect in the Region window;
double-click on the first cross section you want to inspect; the Cross Section editor/view is
activated;
activate the Show/hide last Selected Cross sections
in the Map ribbon (??);
and scroll through the cross sections in the Region window by using the up/down arrow
keys.
54
Deltares
Module D-Flow 1D: All about the modeling process
4.3.13
4.3.13.1
General functions on network objects
Esc key
The Esc key is handy to stop the editing mode (Add ...) and switch to selection mode.
4.3.13.2
Copy and paste network object
4.3.13.3
T
To copy and paste network objects (weirs, pumps, extra resistance, etc.) select the object you
want to copy. Choose Copy from the context menu (right mouse-click). Select a branch you
want to paste the object into by a left-mouse-click. Right-mouse-click the branch to open the
context menu and select Paste. Move the mouse until the cursor is on the desired position
and click the left mouse-button.
Add network object
DR
AF
Network objects (weirs, pumps, extra resistance, etc.) can be added to the network in two
ways:
click on the appropriate button in the Map ribbon (??). Then click on the preferred location
in the network to position the object. The object is snapped to the nearest location on a
branch.
right-mouse-click in the Region window, on the branch, and select Add object. The object
is added at zero chainage. This can be adjusted in the Properties window.
4.3.13.4
Zoom to network object
It is possible to zoom in to network objects by right-mouse-click on the object in
the Attribute Table
the Region window; for Laterals in the Project window
and select Zoom to feature.
To return to the overall view right-mouse-click on the network in Map window and select Zoom
to extend.
4.3.13.5
Selection of multiple network objects
The simplest way is to select
will be high-lighted.
in the Map ribbon (??) and swipe the map. The selection
Another way is as follows:
Press the Esc key
Select the first network object by a left-click on the map
Hold the Ctrl or Shift key while selecting the next one
(the selection will be high-lighted)
While holding the Ctrl or Shift key, a left-click will de-select the network object
Deltares
55
SOBEK 3, User Manual
DR
AF
T
Now the user can delete all or modify one of the properties in the Properties window.
56
Deltares
Module D-Flow 1D: All about the modeling process
DR
AF
T
Types of boundary conditions
Figure 4.34: Example of a network with nodes with or without boundary conditions
D-Flow 1D provides different types of boundary conditions:
No-flow (no boundary condition)
H boundary condition (waterlevel boundary condition, boundary condition of the first kind,
Dirichlet boundary condition)
4.4.1
Boundary conditions
H: constant waterlevel
H(t): waterlevel as a function of time
Q boundary condition (discharge boundary condition, boundary condition of the second
kind, Neumann boundary condition)
4.4
Q: constant discharge
Q(t): discharge as a function of time
Q(h): discharge-waterlevel-relation (rating curve)
In D-Flow 1D, boundary conditions are a property of a Node (section 4.3.2.1). Discharge
boundary conditions can only be applied on Nodes on a single branch on the model boundary,
whereas waterlevel boundary conditions can be applied also on Nodes that connect multiple
branches.
Deltares
57
SOBEK 3, User Manual
Editing boundary conditions
T
The boundary conditions are edited by double-clicking <Boundary Data> in the Project. In
the Central Map the boundary nodes are presented on the map and listed in a table.
DR
AF
4.4.2
Figure 4.35: Boundary nodes in the Central Map
Right-mouse-clicking on one of the nodes in the table and selecting Open View ... opens an
editor. The following types of boundary conditions can be selected:
None
H(t): waterlevel time series
Q(t): discharge time series
Q(h): discharge waterlevel relation table
Q : constant discharge
H : constant waterlevel
By default, each Node (section 4.3.2.1) is a no-flow boundary condition. This means no water
enters or leaves the model.
58
Deltares
Module D-Flow 1D: All about the modeling process
Time series for boundary conditions
DR
AF
T
To generate a time series, right-mouse-click in the Project window on the specific Boundary
Node and select Generate data in series (Figure 4.26).
Figure 4.36: Timeseries on boundary node
Properties of a Timeseries can be adjusted in the Properties window:
Extrapolation type for ..., choose:
Constant (default)
Linear
Periodic
None
Interpolation type for ...
4.4.3
Constant
Linear (default)
It is also possible to use a Timeseries which is available in the <Project>. Chapter section 4.2.5 describes how a Timeseries can be imported and linked to a Boundary node.
Deltares
59
SOBEK 3, User Manual
Simulation results corresponding to discharge boundary conditions
T
4.4.4.1
Remarks on discharge boundary conditions in D-Flow 1D
DR
AF
4.4.4
Figure 4.37: Computational grid of a simple network with a discharge boundary condition
upstream (water flows from right to left).
Figure 4.38: Side-view of computed waterlevels corresponding to the model given in Figure 4.37 (water flows from left to right, discharge boundary condition upstream). The distance between Calculation points is 500 m.
A waterlevel boundary condition is applied on the first Calculation Point (see section 4.9)
next to the Node with the boundary condition. This Calculation Point usually has the same
60
Deltares
Module D-Flow 1D: All about the modeling process
coordinates as the Boundary Condition Node. However, a discharge boundary condition is
not applied on a gridpoint, but on a reach segment (see also section 4.9) because of the
staggered grid numerical scheme (Stelling and Duinmeijer, 2003; Stelling and Verwey, 2006).
So D-Flow 1D sets a discharge boundary condition on the reach segment that is connected
with the Boundary Condition Node. The Calculation Point corresponding with the Boundary
Condition Node is not taken into account within the solution of the equation system, and
consequently no waterlevel result is assigned to this Calculation Point. As an estimation,
the result of the neighboring downstream gridpoint is copied to the Calculation Point at the
Boundary Condition Node (see Figure 4.37). This is physically not correct and has to be
taken into account in the design of the model and the analysis of simulation results:
A Node with a discharge boundary condition should not represent a gauge. Use an Ob-
T
DR
AF
servation Point instead and extend the upstream end of the branch in such a way that the
observation point is located between two gridpoints that are considered in the solution of
the flow equations. In other words: the observation point should not be located within the
first and the second gridpoint that follow the Boundary Condition Node on a branch.
Simulation results of a SOBEK-RE model that has been imported into D-Flow 1D (or
SOBEK 2) will differ from results of the original SOBEK-RE model at nodes with discharge
boundary conditions.
The upstream end of a side-view will always show a horizontal course of the waterlevel
in case of a discharge boundary condition between the two upstream grid points (Figure 4.38). This is physically not correct in most cases.
Results related to Nodes with discharge boundary conditions should not be used to produce rating curves.
The usage of Discharge Boundary Condition Nodes as exchange items in an OpenMIcomposition can produce unexpected results (Becker and Gao, 2012).
Modelers experience the limitations of a discharge boundary condition as a weak point of
D-Flow 1D. Future releases of D-Flow 1D will provide an improved discharge boundary condition.
4.4.4.2
Discharge-waterlevel-relation
In case of a discharge-waterlevel-relation (rating curve, Q(h)), the discharge value is determined with the help of a waterlevel-discharge-relation-table. As input the waterlevel from the
previous time step is used.
It is also possible to model a waterlevel-discharge-relation (h(Q)). To do so, the user must
specify negative discharge values in the waterlevel-discharge-relation-table.
Deltares
61
SOBEK 3, User Manual
Setting the initial conditions
T
4.5.1
Initial conditions
DR
AF
4.5
Figure 4.39: Initial conditions editing in a table (below the Central Map) for branchchainage locations and the corresponding initial value
For a D-Flow 1D model
water depth or waterlevel; and
discharge
can be set as initial conditions. By double-clicking in the Project window on <water flow
model 1d(1)/Input/Initial conditions/initial water depth> the initial conditions are presented (as
a separate layer) in the Central Map and in a table (Figure 4.39).
To define initial conditions with spatial variation, add locations in the table by mouse-clicking
in the Map ribbon. A location can now be added by a mouse-click on the location in the
map. The location is added to the table (Figure 4.39), in which the chainage and value can
be adjusted. Network locations can also be added by directly adding a new line and providing
branch and chainage data in the table.
As soon as a network location on a branch is defined, the default value for the schematization
is overruled for that branch by the locally defined value. When more network locations are
added to the same branch, the values are interpolated linearly between the locations and
extrapolated constant towards the nearest node, see also Figure 4.39.
The initial conditions for discharge can be specified similarly by double-clicking in the Project
window on <flow model 1d(1)/Input/Initial conditions/initial water flow>. Positive discharge
values means water flowing in the direction of the branch, a negative value means water
62
Deltares
Module D-Flow 1D: All about the modeling process
flowing opposite of the defined direction.
Initial conditions from restart
Instead of prescribing initial conditions, it is possible to start a model run from a previously
calculated model state: a restart. A restart state is a complete model state including the values of all the relevant parameters (waterlevels, velocities, discharges, positions of structures,
numerical parameters, etc.) required to reproduce exactly the same simulation results starting
from this restart state as from the original simulation that created the restart state. A model
can only restart from a previous model run for the same model. For restart options see also
section 4.10.
T
Of course, the (restart)states must be available in the project. This can be achieved by selecting <flow model 1d(1)> in the Project window and set <Write restart> on <TRUE> in the
Properties window (as in Figure 4.40).
DR
AF
4.5.2
Figure 4.40: Flow model properties window: How to write restart states
This way, the state will be stored in the Project window on <flow model 1d(1)/Output/States/...>,
see(as in Figure 4.40).
Deltares
63
DR
AF
T
SOBEK 3, User Manual
Figure 4.41: States calculated in previous model run
In order to use this state, the user must select and drag the state to the <flow model 1d(1)/Intput/Initial
conditions/...> in the Project window, and set <Write restart> on <TRUE> in the Properties
window (as in Figure 4.40).
Figure 4.42: Flow model properties window: How to use a restart state
a previous simulation of the same model
64
Deltares
Module D-Flow 1D: All about the modeling process
a simulation in the same project
The only restriction is that the network has to be the same.
4.6
4.6.1
Roughness
Introduction
DR
AF
a constant value
spatially varying
a function of waterlevel or discharge.
T
The roughness of the bed is defined for the entire width of the branch. The branch can be
divided in separate Sections (roughness) with different roughness characteristics, for example
main channel (summerbed) and left and right bank (winterbed). The user is free to choose
names for the roughness-sections. An exception is the symmetrical Cross Sections ZW (section 4.3.12.4). If this type of Cross Section is used, the roughness-sections have pre-defined
names: Main, FloodPlain1 and FloodPlain2. For each roughness-section on a branch the
roughness can be specified as
The roughness values themselves can vary along the branch, so roughness can be allocated
for any number of locations along a branch.
The roughness values per roughness-section can be edited as a separate model feature. This
has the advantage that the roughness for all locations is directly visible in one table or map.
This gives the user a good overview of the roughness in the network. The roughness is also
easier to edit.
4.6.2
Defining roughness
Figure 4.43: Roughness editor for a model of the Dutch part of the river Meuse. On the
left the roughness table with Branch, Chainage, Function, Roughness Type,
Value and Unit (automatically set according to the Roughness Type); on the
right the graphical representation of the roughness-table content.
Deltares
65
T
SOBEK 3, User Manual
Figure 4.44: Setting of roughness-sections in the Region window
DR
AF
Defining roughness is a three-step process:
1 Define the roughness-sections, e.g. main, left bank, right bank and so on. In case of Cross
Sections ZW this step can be omitted as the names are pre-defined.
2 Define the geometry of the roughness-sections in the cross-section editor (section 4.3.12,
Figure 4.27, Figure 4.28, Figure 4.30).
3 Set the roughness-type and -values in the roughness editor which is accessible by doubleclicking Roughness in the Project window. By adding locations on the branches, the
roughness can be specified varying over the network as shown in Figure 4.43.
A roughness-section is added in the Region window by a right-mouse-click on
and choosing Add Section Type (Figure 4.44). The roughness-section is added to the list and
can be renamed by a double mouse-click on the specific roughness-section or by pressing
the key “F2”.
66
Deltares
T
Module D-Flow 1D: All about the modeling process
DR
AF
Figure 4.45: Cross section editor for an XYZ Cross Section with three Sections (roughness). The roughness-section ‘left bank’ is selected in the table and highlighted in purple.
For each Cross Section the roughness-sections need to be set in the cross-section editor
(section 4.3.12, Figure 4.27, Figure 4.28, Figure 4.30) in the table (Cross Section ZW: values
can be specified for “Main”, “Floodplain1” and “Floodplain2”). Fill in the Start and the End
columns with Y -values (or Y 0 -values for Cross Sections XYZ) and chose the Roughness in
the drop-down menu. The roughness-sections are visualized as blocks beneath the graphical
representation of the cross-section (Figure 4.45).
The list of roughness sections is also visible in the Project window. A double click on a specific
roughness section opens the roughness editor for this section (Figure 4.43) with a roughness
table and its graphical representation. The columns Branch and Chainage in the table define
the location in the network. With the Add Network Location tool
from the Menu bar,
locations can be added to the table by a mouse click in the map. These locations can be
moved to a precise location by adjusting the chainage value in the table. Within a branch the
roughness values are interpolated between the specified network locations. If no locations
are specified for a branch, the default value is used for the entire branch. The roughness can
be defined as
constant
a function of water level h
a function of discharge Q.
Deltares
67
DR
AF
T
SOBEK 3, User Manual
Figure 4.46: Function table for roughness as a function of discharge and the graphical
representation of the table content
The following rouhgness parameters (Roughness Type) are available:
Chézy
Strickler ks
Strickler kn
Manning
White & Colebrook
Bos & Bijkerk
The choice of Function Type in the roughness table (in our case “FunctionOfQ” is valid for
the whole branch, so the corresponding drop-down menu is only accessable for a chainage of
0 m. In case of a constant or spatially varying roughness, the value is set in the column Value.
If the roughness depends on water level or discharge, the corresponding function has to be
specified in a function table (left mouse-click on the corresponding field in the last column
of the roughness table
). For a branch in the Meuse model (Figure 4.43) such
a function table is given with Figure 4.46. Here the roughness is defined as a function of
discharge Q for the whole branch. The first column in the function table Figure 4.46 contains
the discharge levels, the remaining columns refer to the chainage values specified in the
roughness-table. If no locations are defined for a branch, the model wide value and type are
used, visible and editable in the Properties window after selecting the roughness coverage.
68
Deltares
Module D-Flow 1D: All about the modeling process
4.6.3
Import and export roughness from/to csv-file
To export roughness definitions right-mouse-click in the Project window on <Roughness>
and select Export. . . . Give a file name in the file selection-window that pops up. In the same
way the values for a single roughness-section can be exported as well. Import works the
same way. The roughnesses can then be modified using another application and imported
again.
Wind
Wind friction
Wind shielding
T
In channel systems with long stretches of narrow channels and/or large open water surfaces
shear stress induced by wind on the water surface can have an impact on the water movement. This leads to locally higher or lower waterlevels than for a situation without wind. In
D-Flow 1D there two effects of wind can be taken into account:
DR
AF
4.7
The wind friction depends on the wind direction and velocity. In D-Flow 1D a spatially uniform,
but temporarily varying wind velocity field can be applied. The wind field can be edited after
double-clicking in the Project window on <water flow model 1d(1)/Input/Initial conditions/wind>.
A time series can be generated by a right-mouse-click on <wind> in the Project window and
choosing Generate data in series (see also Figure 4.26).
Wind shielding is a geometrical effect; parts of a river may be in the lee and in practice feel
only part of the wind or no wind at all. Wind shielding is modelled in D-Flow 1D as a factor
which determines the fraction of the total wind field actually impacting the channel. The values
range from 1 (no shielding) to 0 (complete shielding). Wind shielding in D-Flow 1D is spatially
varying, but uniform in time.
Deltares
69
DR
AF
T
SOBEK 3, User Manual
Figure 4.47: Wind shielding (factors) presented in the Central Map and the table for editing
The default factor can be adjusted in the Properties window when in the Project window
<flow model 1d(1)/Input/Initial conditions/wind shielding> is selected. The wind shielding
editor (Figure 4.47) opens on double-clicking in the Project window on <water flow model
1d(1)/Input/Initial conditions/wind shielding>.
Network locations can be added to the table with the help of the Add Network Location tool
in the Map ribbon (??). Move the mouse to a location in the map of the wind shielding
editor and left-click. The location is added to the table as a value pair of Branch and Chainage.
Adjust the value in the table if necessary. Network locations can also be added by adding a
new line in the table.
If a network location on a branch is defined, the default value for wind shielding is overruled. When more network locations are added to the same branch, the values are interpolated linearly between the locations, or values are extrapolated constantly towards the nearest
node. The results of interpolation or extrapolation are visualized in the wind shield editor (Figure 4.47). If no wind data (friction or shielding) is specified, D-Flow 1D assumes no influence
of wind on the water flow.
70
Deltares
Module D-Flow 1D: All about the modeling process
Salt water intrusion
T
Saltwater intrusion means the movement of a salt water wedge into an estuary following
Thatcher and Harleman (1972). The term “saltwater intrusion” is also used for the movement of saline water into freshwater aquifers. “Intrusion” is a geological term used for the
process of liquids into hard rock (Wikipedia, 2010). Salt transport in estuaries and tidal rivers
can be considered as transport of conservative substance in water. The transport of salt is
described by the advection-diffusion equation for the salt concentration or the chloride concentration. In this way, density differences are introduced that have to be accounted for in
the momentum equation of the water flow module. The flow field as computed by the flow
model will be used again in the advection-diffusion equation of salt, and so on. The water flow
module is therefore coupled with the salt intrusion module by the density and the flow field
(RIZA, 2005).
The process of salt water intrusion can be added to the D-Flow 1D model by selecting <water
flow model 1d (1)> in the Project window, and setting Use salinity in the Properties window
to True, (Figure 4.48, see also section 4.10). The property Use salt in calculation has been
implemented to leave out the salt water intrusion processes in the flow simulation without
losing the salt related data in the schematization. When Use salt is set to False, all existing
salt data in the schematization is deleted after having warned the user with the help of a
message box.
DR
AF
4.8
Deltares
71
DR
AF
T
SOBEK 3, User Manual
Figure 4.48: Addition of salt in a flow model in the Properties window
When salt water intrusion processes are added to the model, the Project window shows two
new components (Figure 4.49):
Initial salinity concentration
Dispersion coefficient
72
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.49: Project window after setting Use salinity to “True”
By double-clicking <Initial salinity concentration> in the Project window the initial salinity
conditions editor is opened. Similarly to other initial conditions a default value for initial salinity
can be set in the Properties window when selecting <Initial salinity concentration> in the
Project window. To define local initial salinity concentrations, add locations to the network
by mouse-clicking the Add Network Location
in the Map ribbon. A location can now be
added by a mouse-click on the location in the map in the initial salinity concentration editor.
The location is added to the table, in which the chainage and branch value can be adjusted.
Network locations can also be added by directly adding a new line and providing branch and
chainage information in the table in the initial salinity concentration editor.
The dispersion coefficient can be spatially uniform or spatially varying and is constant in time.
The dispersion coefficient can be edited similarly to the initial salinity concentration. An advanced option is the use of the Thatcher-Harleman dispersion formulation. In this case a
time-dependant dispersion is evaluated and two tuning coefficients are important. This can
be activated as follows:
double-click on <Dispersion coefficient> in the Project window
select Initial Dispersion View
check Use Thatcher-Harleman
now the following coefficients appear (in stead of Dispersion Coefficient):
specify F1
specify F3
Deltares
73
DR
AF
T
SOBEK 3, User Manual
Figure 4.50: The use of Thatcher-Harleman dispersion formulation
When salt water intrusion processes are taken into account in the D-Flow 1D model, salt is
also added in the boundary node editor. By double-clicking on the specific boundary node
in the Project window, the boundary node editor is opened in which the user can select
the option Edit salinity data. Figure 4.51 shows the resulting screen. The user specifies a
salt concentration at the boundary either as a constant or as a function of time. For salinity
concentrations as a function of time, a time series can be generated by adding dates to the
table. At tidal sea boundaries, the water will be alternately flowing out of the model and
into the model. The Thatcher-Harleman time lag defines a transition period in seconds for the
boundary condition when the condition changes from low tide to high tide the model (Thatcher
and Harleman, 1972; RIZA, 2005).
74
Deltares
DR
AF
T
Module D-Flow 1D: All about the modeling process
Figure 4.51: Boundary node editor for salinity
4.9
Computational grid
D-Flow 1D uses a staggered grid for the numerical solution of the flow equations (Deltares,
2013). The computational grid is not part of the D-Flow 1D network, but a separate layer
which can be opened and viewed in a map by double-clicking in the Project window on
<computational grid>.
Deltares
75
DR
AF
T
SOBEK 3, User Manual
Figure 4.52: Generate Computational Grid window
To generate a computational grid for a network right-mouse-click <computational grid> in the
Project window and select Generate calculation grid locations. The grid generator window
(Figure 4.52) appears. By default, the grid is generated for the entire network. If one or
multiple Branches are selected in the computational grid view and “Selected branches” is
activated in the computational grid editor, a computational grid is generated only for selected
Branches. There are two general options for the grid generation:
Generate new calculation points. This option removes all existing calculation points. A
completely new grid is generated.
Use existing calculation points. With this option the existing calculation points are reused
for the branch where they are already present.
For the positioning of the calculation points the following options are available:
None. This option removes the grid from a branch,
Prefered length. This option defines the prefered distance between calculation points.
Special locations.
76
Cross Section. With this option D-Flow 1D generates also a calculation point on each
Cross Section the Network/Branch
Lateral Sources. A grid point is generated on the location of a Lateral Source. As
the continuity equation is computed for grid points, it can be advantageous for water
balance studies or water quality modeling studies to set a grid point on Lateral Source
locations.
Structures. When a structure is present on a reach segment between calculation
Deltares
Module D-Flow 1D: All about the modeling process
points, the characteristics of the structure are used for the entire segment. If this option
is switched on, D-Flow 1D generates calculation points upstream and downstream a
structure at a defined distance from the structure to restrict the characteristics of the
structure to a specified region (note that this distance should not be too small for
stability reasons).
T
D-Flow 1D spreads the calculation points uniformly over the branch. If the length of the branch
is not equal to multiple the preferred distance (for example, the length of the branch is 990 m
and the preferred distance is 100 m), D-Flow 1D generates a grid which optimizes the number
of calculation points and their distance as close to the preferred distance as possible (in the
example D-Flow 1D generates calculation points with a uniform distance of 99 m instead of
nine times a distance of 100 m and once a distance of 90 m).
DR
AF
With the grid generator functionality it is easy to experiment with different grids to find a suitable one which is fine enough, but not too computationally expensive. A grid can be considered to be fine enough if the simulation results do not change significantly if the grid is further
refined. A starting point for the distance between grid points is the width of the cross sections.
Keep into account that:
the distance between calculation points should not be too large to ensure sufficient accuracy,
the distance between calculation points should not be too small for stability and calculation
time. By default, the smallest possible distance in the numerical scheme is set to 10 m,
the distance between the calculation points may be non-equidistant.
Figure 4.53: Table and map view of the computational grid (note that only waterlevel
points are shown in this view)
To visualize the computational grid double-click on <computational grid> in the Project window. Figure 4.53 shows the editing window of a computational grid. Note that this layer shows
only the waterlevel-points and not the velocity points of the computational grid. In the table the
Deltares
77
SOBEK 3, User Manual
Branch, the chainage, the gridpoint ID and the grid point type of waterlevel points are given.
A grid point type of zero represents a non-fixed grid point, one means fixed grid point. To
change the grid point type, select a calculation points and select Fixed gridpoint in the context
menu. The grid point type can also be changed in the table by editing the field in the Grid
point type column. Fixed calculation points are not affected when the grid is redefined and
are shown on the map in a different color. To add Calculation points use the Add Network
Location tool
4.10
Model properties
Introduction
T
4.10.1
in the Map ribbon (??) and click on the preferred locations in the map.
4.10.2
General
DR
AF
When a flow model in the Project window is selected, in the Properties window the modelwide settings can be specified. These settings are supplied to the calculation core at Run
model. The parameters are divided in different categories which are discussed below. The
parameters mentioned in this section are present for any D-Flow 1D model. Not all parameters are elaborated here. In addition, there are some parameters which are only used in the
Grafical User Interface.
In this category the user can only set the Name of the flow model.
4.10.3
Initial conditions
As initial conditions the user can specify waterlevel or water depth. For both options a global
value can be specified. For lowland areas the waterlevel can be an appropriate option, for
hilly modeling areas the water depth option can be the option of choice.
The use of previously computed simulation results as initial condition (Restart) is described in
section 4.5.2.
4.10.4
4.10.4.1
Model settings
Roughness for tidal flow
With regard to tidal / reverse flow, the user can set the following parameters:
Use reverse roughness, default: False
Use reverse roughness in calculation, default: False
78
Deltares
Module D-Flow 1D: All about the modeling process
4.10.4.2
Salt water intrusion
With regard to salt water intrusion, the user can set the following parameters:
UseSalinity, default: false
UseSalinityInCalculation, default: false
Use Thatcher Harleman, default: False
and under /buttonRun parameters / Model parameters:
[65] DiffusionAtBoundaries, default: false
[69] DispMaxFactor, default: 0.45
T
The difference between UseSalinity and UseSalinityInCalculation is that the first one
is related to salinity data and the second one related to the flow simulation. If UseSalinity
is true and UseSalinityInCalculation is false, the simulation is run without salt, but the
salt-related data in the model are still present, like initial salinity concentration. A
next simulation can then be performed with salt without having to set all data again.
DR
AF
DiffusionAtBoundaries makes it possible to switch the diffusion term at boundaries on
or off. For modeling of salinity a so-called advection-diffusion equation is applied. At open
boundaries the user has the possibility to switch on or off the diffusion term. The default
option is that the diffusion is switched off at open boundaries.
For modeling of salinity SOBEK uses an explicit numerical method. This requires time step
limitations in order to ensure stability. For the dispersion term this is of the form
∆t · D
≤ LD
2∆xi 2
(4.1)
with the dispersion coefficient D and LD the dispersion limit. ∆xi is the mesh size (of node i)
and ∆t denotes the time step. In SOBEK 3 a value of 0.45 is applied for LD , which is slightly
smaller than the theoretical maximum of 0.5. It is advised not to change this model parameter.
It is foreseen that in a next release of SOBEK 3 this model parameter will be removed, since
an implicit scheme will be implemented.
4.10.5
Output parameters
In this category the user can set the Model output time step. The user can specify which
type(s) of output in section 4.11.
4.10.6
4.10.6.1
Run parameters
Simulation period and timestep
The user specifies:
StartTime: start point in time of simulation period in date format [yyyy-mm-dd hh:mm:ss].
Default: yesterday, 00:00:00 h.
StopTime: end of the simulation period in date format [yyyy-mm-dd hh:mm:ss]. Default:
today, 00:00:00 h.
TimeStep: spatial discretization for the simulation [dd hh:mm:ss]. Default: 0 d, 01:00:00 h.
Deltares
79
SOBEK 3, User Manual
4.10.6.2
Restart and save State
In this category the user writes and uses states to restart a simulation:
Use restart: use the State stored in the <Model / Input / Initial conditions / state>. The
user will have to copy or link the State from a previous model run (from <Model / Output>
to <Model / Input ...>)
Use save state time range: copy the StartTime from the saved state
Write restart: write the state and store in the <Model / Output>
Model parameters
T
Here, the following numerical parameters can be specified. Some of these are explained in
detail in the following sections.
[31] AccelerationTermFactor: Factor on 1D acceleration term
0.0 and 1.0, default: 1.0
∂U
∂t ,
can vary between
DR
AF
[32] AccurateVersusSpeed: Accuracy factor, default: 3
[33] CourantNumber: Maximum Courant number, default: 1.0
[35] EpsilonValueVolume: Convergence criterion for water volume balance, default:
0.0001 m3
[36]
[38]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
EpsilonValueWaterDepth: Convergence criterion for water depth, default: 0.0001 m
MaxIterations: Maximum number of iterations, default: 8
MinimumSurfaceinNode: Minimum surface in node, default: 0.1 m2
MinimumLength: Minimum branch segment length, default: 1.0 m
RelaxationFactor: Relaxation factor, default: 1.0
Rho: Density of freshwater, default: 1000 kg/m3
StructureInertiaDampingFactor: Structure inertia damping factor, default: 1.0
Theta: Theta-value, default: 1
ThresholdValueFlooding: Threshold water depth for flooding of channels, de-
fault: 0.01 m
[47] ThresholdValueFloodingFLS: Threshold water depth for flooding of land surface,
default: 0.001 m
[48] UseTimeStepReducerStructures: Use timestep reduction on structures (0=false,
1=true), default: 0
[49] ExtraResistanceGeneralStructure: Extra resistance for general structure, default: 0.0
[51] NoNegativeQlatWhenThereIsNoWater: Limit lateral outflow to the water available
in the channel, default: true
[52] TransitionHeightSD: Transition height for summerdikes, default: 0.5 m
parameters related to quasi steady state mode:
[53]
[54]
[55]
[56]
[57]
[58]
ComputeSteadyState
Dtsteady
EpsMaxU
Ntendcontrolsteady
Ntintcontrolsteady
Ntmaxsteady
parameters for debugging a model:
4.10.6.3
80
[59] Debug, default: false
[62] DebugTime
Deltares
Module D-Flow 1D: All about the modeling process
numerical parameters, read more in section 4.10.6.12:
[65] Iadvec1D: Advection Type in 1-dimensional flow, default: 1
[66] Limtyphu1D: Limiter type for estimating flow area at velocity point in 1D flow,
default: 1
[67] Momdilution1D: Advection control volume based upon flow area or total area
in 1D links, default: 1
4.10.6.4
Structure Inertia Damping Factor
T
The structure equations contain an inertia term. This inertia term acts as a kind of numerical
damping. This is done to avoid numerical oscillations in case of unsteady flow conditions. The
numerical parameter ’factor for structure dynamics’ is a factor applied to this inertia term. As
default value for [44] StructureInertiaDampingFactor 1.0 is suggested. Note that for
steady flow conditions the inertia term is set to zero, because in this case the Structure Inertia
Damping Factor is not taken into account.
DR
AF
The structure inertia damping factor is applied for the River weir, the Advanced weir, the General structure and the Database structure both as single structure or member of a composite
structure. In the linearization of the concerning structure equation a term
α
∂U
∂t
(4.2)
is added, where
α
U
t
structure inertia damping factor [-];
flow velocity [m/s]; and
computational time [s].
The structure inertia damping factor can be used for avoiding instabilities during computation.
4.10.6.5
Quasi steady-state
D-Flow 1D can run a quasi steady-state simulation mode. This means, D-Flow 1D solves the
flow equations for each time step in the simulation period repeatedly until a steady-state is
reached before continuing with the next time step (see Becker and Prinsen, 2010). Because
of these iterations, for small time steps the quasi steady-state mode will be computational
more expensive than a transient simulation. In order to model for example seasonal steady
states, it can make sense to simulate a whole year with 4 quasi steady-state time steps, one
time step for each season. In general it makes sense to apply the quasi steady-state mode if
the signal (i.e. the boundary condition) plays on a time scale which is larger than the time the
signal needs to reach the opposite end of the modeling area. This can be the case for lowflow conditions, for example. If the dynamics of the boundary conditions play on a smaller time
scale than the boundary condition signal needs to reach the other end of the modeling area
(e.g. high-water scenarios, tidal waves), a transient simulation should be preferred against the
quasi steady-state simulation (Becker and Prinsen, 2010).
To run a simulation in quasi steady-state mode, set the following Model paramaters (see
Becker and Prinsen, 2010, for details):
[53] ComputeSteadyState: switch for quasi steady-state simulation mode (“True” for
quasi steady-state simulation mode), default: “False”
Deltares
81
SOBEK 3, User Manual
[54] DtSteady: time step for quasi steady-state simulation [seconds], default: 7200 s
[55] EpsMaxU: a convergence criterium to determine that steady-state conditions have
been reached based on the velocity difference, default: 1 · 10-6 m/s
[56] Ntendcontrolsteady and [57] Ntintcontrolsteady define how often control
is applied during the iterations. Default values: [56] Ntendcontrolsteady = 200,
[57] Ntintcontrolsteady = 20
[58] Ntmaxsteady: the maximum number of iterations for one quasi steady-state time
step, default: 1500
4.10.6.6
Extra resistance for general structure
DR
AF
ρ2 L · W2 · U2 2
or
g
C2
T
A default value is defined for the so called extra resistance coefficient of the General structure
type, both as a single structure and as a member of a structure: [49] ExtraResistanceGeneralStructure
(default: 0.0). This default value can be overruled for each individual General structure type.
The so called extra resistance refers to a bed shear stress force, that is accounted for in the
impuls balance, that is solved in case of drowned gate flow or drowned weir flow. The bed
shear stress force reads
λρ2 · W2 · u2 2
where:
λ
L
g
C
r2
U
W2
.
4.10.6.7
=
(4.3)
(4.4)
L·g
,
C2
extra resistance coefficient;
length of hydraulic jump behind the structure in m;
acceleration due to gravity in m2 /s;
Chézy coefficient in m1/2 /s;
density of water in hydraulic jump in kg/m3 ;
downstream flow velocity in m/s; and
downstream structure width in m
Summerdike
For summerdikes, the user can set the transition height ([52] TransitionHeightSD, default
0.5 m), see also section 4.3.12.
4.10.6.8
Advanced options
The user can set the following advanced parameters:
[41] MinimumLength: Minimum branch segment length, default: 1.0 m
[38] MaxIterations: Maximum number of iterations, default: 8
[51] NoNegativeQlatWhenThereIsNoWater: Limit lateral outflow to the water available
in the channel, default: true
82
Deltares
Module D-Flow 1D: All about the modeling process
4.10.6.9
Volumes based on waterlevels or discharges
In SOBEK there are two options for computing the volume of a calculation point (or reach
segment) at a specific point-in-time, viz:
Volumes based on waterlevels (parameter value is 0)
Volumes based on discharges (parameter value is 1)
4.10.6.10
DR
AF
T
If the option ’Volumes based on waterlevels’ is selected, this means that the volume at each
calculation point (or each segment) follows from the computed waterlevel and its corresponding cross sectional cross section. If the option ’Volumes based on discharges’ is selected, this
means that the volume at each calculation point (or reach segment) is the summation of its
volume in the previous time-step and the resulting net inflow during the computational timestep. In Water Quality computations especially use is made of volumes and discharges. By
choosing the option ’Volumes computed based on discharges’ a more coherent set of volumes
and discharges is obtained, than in case the option ’Volumes computed based on waterlevels’
is selected.
Reduction of timestep on large lateral flow
In case the user defines a value for ’Reduction of time step on large lateral inflow’ equal
to 1, the computational time step will be reduced in such a way that the maximum lateral
inflow volume is not more than the volume stored in the corresponding computational point.
In addition the user can define a minimum time step for the ’reduction of time step on large
lateral inflow’ procedure by defining a value for the parameter ’minimum time step in time
step reduction on large lateral flow’. Whether the ’Time step reduction on large inflow’ is
true or false, for lateral outflow always the above procedure applying the actual value for the
parameter ’minimum time step in time step reduction on large lateral flow’ is used.
4.10.6.11
Use timestep reduction on structure
In case the user defines a value for [48] UseTimeStepReducerStructures equal to 1, at
the point-in-time of the wetting of the crest of a structure (i.e. for weirs and orifices only) a time
step reduction will be applied during a time-span equal to two times the user defined time step.
This functionality was implemented to avoid oscillation in specific Urban schematisations with
sharp inflow hydrographs, it can be applied in Rural schematisations as well. Unnecessary
use of this option might result in a longer computational time needed.
Deltares
83
SOBEK 3, User Manual
4.10.6.12
Parameter set for lowland rivers
Three numerical parameters are specially suited for lowland rivers with strong contraction
and/or expansion. The third is new in SOBEK 3.
[65] Iadvec1D: This parameter determines the way the advection term in the De Saint
1: Conservation of Momentum
2: Balanced Average of Conservation of Momentum and Conservation of Energy in
Contraction and Expansion
3: Balanced Average of Conservation of Momentum and Conservation of Energy in
Contraction Only
4: Balanced Average of Conservation of Momentum and Conservation of Energy in
Expansion Only
5: Balanced Average of Conservation of Momentum and Conservation of Energy but
no Contraction and Expansion Losses
T
Venant equation is implemented, default: 1 :
[66] Limtyphu1D: This parameter determines the estimation of the waterlevel at the
DR
AF
velocity points to calculate the continuity equation, default: 1 :
1: Upwind
2: Central in Cross-sections
3: Central in Water levels
[67] Momdilution1D: Advection control volume based upon flow area or total area in
1D links, default: 1 :
1: Total area
2: Flow area with account for storage sink term
3: Flow area
For lowland rivers choose:
[65] Iadvec1D: 2
[66] Limtyphu1D: 2
[67] Momdilution1D: 1
4.10.7
Default bed roughness
The (factory) defaults for the roughness type and value are:
Roughness type (Default: Chézy)
Default roughness value (Default for Chézy: 45 m1/2 /s)
The user can overrule this default value by defining the roughness locally, see section 4.6.
The options for roughness types and their corresponding default values are given in table 4.1.
84
Deltares
Module D-Flow 1D: All about the modeling process
Table 4.1: Options for roughness types and default values
Chézy C
Manning Mn
Strickler Ks
Strickler Kn
Bos & Bijkerk γ
White & Colebrook Kn
Output
45
0,03
33
0,2
33,8
0,2
Unit
m1/2 /s
s/m1/3
m1/3 /s
m
m
Before running a simulation, the user can set which output is required by selecting <output>
in the Project window under <water flow model 1d (1)>. The Properties window then looks
like Figure 4.54.
DR
AF
4.11
Default value
T
Roughness type
Deltares
85
DR
AF
T
SOBEK 3, User Manual
Figure 4.54: Set output in the Properties window
The list in the Properties window contains all possible parameters for which simulation results
can be generated. For each parameter, the user can choose between the following types of
output:
Maximum: the maximum value during the output timestep
Minimum: the minimum value during the output timestep
Average: the average value during the output timestep
86
Deltares
Module D-Flow 1D: All about the modeling process
Current: the values at the precise timestep
In addition, two output timesteps can be set:
Gridpoints: for gridpoints and reach segments
Structures: for structures, lateral sources, retentions and observation points
Water depth
Waterlevel
Water volume
Total area
Total width
Density
Salt concentration
Salt dispersion
DR
AF
T
The user can choose the following parameters on gridpoint locations:
For reach segments, the following parameters are available:
Chézy values
Conveyance
Discharge
Flow area
Froude number
Hydraulic radius
Subsection parameters
Velocity
Waterlevel gradient
For structures, the following parameters are available:
Crest level
Crest width
Discharge
Flow area
Gate lower edge level
Head difference
Opening height
Pressure difference
Valve opening
Velocity
Waterlevel at crest
Waterlevel down
Waterlevel up
For lateral sources, the following parameters are available:
Discharge
Waterlevel
Deltares
87
SOBEK 3, User Manual
For observation points, the following parameters are available:
Discharge
Velocity
Water depth
Waterlevel
For retentions, the following parameters are available:
Volume
Waterlevel
4.12
DR
AF
Negative depth
Number of iterations
Timestep estimation
T
For simulation info, the following parameters are available:
Validation
As a final step in the modeling process, the user can activate the validation tool, by rightmouse-click in the Project window on <water flow model 1D> and selecting Validate.... A
validation report is presented in the central window (Figure 4.55). This example shows the
Figure 4.55: Validation Report: example
validation report for a simple flow model where the computational grid has not yet been defined. The user can simply double-click to open the appropriate editor. This way the report
serves as a todo-list.
The validation tool checks all that is required for a model run. In other words: a validated
model will run!
88
Deltares
5 Module D-Flow 1D: Simulation and model output
When the schematization is complete, the model is ready for a simulation. A simulation is
run by right-mouse clicking the flow model in the Project window and selecting run model.
Alternatively, by clicking one of the buttons in the Home ribbon:
DR
AF
T
During the simulation a progress bar appears and simulation messages are shown in the
Messages window. A logfile (Run report) is added to the model output in the Project window,
which can also be exported. In this report all the schematization and simulation messages are
logged. During the simulation, output is generated by the model. Besides physical quantities
related to flow, such as velocity or water level, also simulation information is provided which
contains information on the accuracy and numerical behaviour of the simulation. The output
is stored in coverages, which are added to the model output in the Project window, see also
Figure 5.1.
In addition, States are stored for later use as Restart. Of course, the user has to request to
Write restart in the Properties window before running the model.
If the network, computational grid or model parameters change, the output is no longer valid
and is deleted. Output in a model is therefore always consistent with the model.
Deltares
89
DR
AF
T
SOBEK 3, User Manual
Figure 5.1: Output in the Project window
5.1
Simulation information
The simulation information is divided in spatial and non-spatial information. The spatial information is generated in output coverages and can be visualized in maps, graphs and tables,
just like other output parameters such as waterlevels. The non-spatial information is saved in
a textfile and can be found in the Project window under model output.
Non-spatial information consists of
Version information of the plug-ins (modules) used
Total calculation time
List of numerical parameters and the values used
Smallest and largest timestep
Water balance components
Balance error
Initial conditions
This information can be used to assess the model performance and accuracy. It can also be
90
Deltares
Module D-Flow 1D: Simulation and model output
used to solve problems in the schematization.
The spatial simulation information consists of
negative depth
Timestep estimation
Number of iterations
This information can be used to remove errors from the schematization or improve the performance of the model.
T
Results in the Map
By double-clicking a specific output parameter in the Project window the results for that parameter are presented (as a separate layer) in the map with the network, see also Figure 5.2.
The map shows for all available calculation points (for that parameter) the value of the specific output parameter for a specific timestep which can be adjusted in the Time Navigator
window. By sliding the red bar in the Time Navigator, the user can navigate through the
results in time. For each map there is a separate Time Navigator, so that it is possible to
view different timeslices of several parameters simultaneously by docking the map windows
next to each other.
DR
AF
5.2
Figure 5.2: Map results of discharge
For each map it is possible to add shapefiles as background map and display or hide (parts
of) the network by (de)selecting the appropriate layers in the Map window. The Map window
can also be used to change the symbols in the map for each layer. By double-clicking on a
layer the Layer properties editor opens in which colorscales, symbol sizes, legend classes
and symbol style can be adjusted, Figure 5.3.
Deltares
91
DR
AF
T
SOBEK 3, User Manual
Figure 5.3: Layer properties editor
Alternatively, a map can be customised by adding a new map. A new map may be opened by
right-mouse clicking on <Project> in the Project window and selecting New item and Map.
parameters can be dragged from the Project window into the map. In this way maps can
be customised by the user. It is possible to combine several parameters from one model, or
parameters from different models, add shapefiles, show (parts of) the network etc. A resulting
map with both water level and discharge is shown in Figure 5.4.
92
Deltares
DR
AF
T
Module D-Flow 1D: Simulation and model output
Figure 5.4: Customised map
5.3
Results in a Graph
Simulation results can also be shown in graphs. By double-clicking on an output parameter
it will be presented in the map. In the map one or more calculation points can be selected.
By clicking
in the Menu bar, Figure 5.5 appears. The user can now select one or more
parameters, which are then displayed in a graph, Figure 5.6.
Figure 5.5: Select parameter for graphical representation
Deltares
93
DR
AF
T
SOBEK 3, User Manual
Figure 5.6: Time results of water level for 3 lcoations along the branch
5.4
Results in a Table
Next to graphs and maps are tables with the actual values of the parameters shown. For map
representations of results, the table shows all locations for one timestep, see also Figure 5.2.
For graphical representations the table shows the selected locations and parameters for the
entire simulation period, see also Figure 5.6.
5.5
Sideviews
To view simulation results along branches a sideview can be opened. First, the user needs to
specify a route.
5.5.1
Routes
In the map with the network, the user can specify a route by selecting
A new (empty) route will be added
stored.
in the Map ribbon.
in the Network window - several routes can be
Also, the
button is activated. By clicking in the map network locations can be
added to the route. The first location marks the starting point of the route, each click on the
map marks either an intermediate point along the route or the end point (in case that network
location was the last one added). The route can be finalized by pressing Esc. The routes can
be altered by moving the network locations by pressing the move-features button
in the Edit section of the Map ribbon and moving the network locations along a route. At any
time new network locations can be added to a selected route by clicking on
.
Note that new network locations are always added to the route, it is not possible to add a
network location halfway. It is possible to add a network location and then move the locations
94
Deltares
Module D-Flow 1D: Simulation and model output
according to the users wishes. Each route has a chainage starting from 0 at the starting
network location. Figure 5.7 shows an example with three network routes.
Intermediate points can be used when there are more options to connect two locations. Without using intermediate locations SOBEK will choose the shortest connection between two
locations as route. By adding intermediate locations, the user can specify alternative routes,
see Figure 5.8.
By right-mouse-click on the route in the Network window, the user can:
T
Open or Open with ... to view the route in the map or in Side View (see next paragraph)
Zoom to feature
Rename
Delete or
inspect Properties
DR
AF
Figure 5.7: Example of 3 network routes shown in the network with different colours
Figure 5.8: Example of the use of intermediate locations to specify routes
5.5.2
Results in Sideview
The user can select a route in the Region window and open a sideview by a mouse-click
on
in the Map ribbon. A sideview always shows the waterlevels, structures and cross
sections. The user can add all available output parameters (from any model run with the same
Deltares
95
SOBEK 3, User Manual
DR
AF
T
network route) and the computational grid, the initial conditions and wind. In this way several
parameters can be viewed simultaneously, see Figure 5.9 for an example with waterlevel and
discharge. Using the Time Navigator, the user can navigate through the results in time.
Figure 5.9: Example of sideview with Time Navigator
5.6
Export
Output data can be exported by by right-mouse clicking one of the model output parameters
in the Project window and selecting export.... The data can be exported in two manners:
Coverage file exporter (NetCDF-format)
FEWS-PI Longitudinal Profiles (FEWS-PI-format)
5.7
Case analysis
Simulation results can be analysed with the Case Analysis (tool) View. This can be activated
by clicking
in the Tools ribbon.
The Case Analysis window pops up (Figure 5.10).
96
Deltares
DR
AF
T
Module D-Flow 1D: Simulation and model output
Figure 5.10: Example of Case analysis
The user can select one of the available results and one of the following Operation(s):
Mean, resulting in the mean value of the simulation period
Min, resulting in the minimum value of the simulation period
Max, resulting in the maximum value of the simulation period
- for the following Operations the user must select a second result or initial conditions:
Add
Substract
Abs(olute) Difference
Figure 5.10 shows the result of a Substraction.
Deltares
97
DR
AF
T
SOBEK 3, User Manual
98
Deltares
6 Module D-Flow 1D: Morphology and Sediment Transport
Introduction
Morphodynamic processes and sediment transport can be simulated with SOBEK 3 as part of
the D-Flow 1D module. At the moment Delta Shell (as User Interface) has only limited support
for morphology: which means that most pre- and post-processing must be done outside Delta
Shell or with the help of Python-scripting.
T
Morphology is activated in the Properties window of <water flow 1d (1)>, see Figure Figure 6.1. The input files must be generated separately, as described in section 6.2. Morphological output cannot be inspected with Delta Shell, but other tools are available, as described
in section 6.3.
A morphodynamic run can be activated in the Properties window after selecting <water flow
1d (1)>, as depicted in Figure 6.1.
DR
AF
6.1
Figure 6.1: How to simulate morfology together with a D-Flow 1D simulation
Deltares
99
SOBEK 3, User Manual
6.2
Input files
Two input files are minimally required for a simulation:
The sediment input file (<∗.sed>) contains the characteristics of all sediment fractions.
The morphological input file (<∗.mor>) contains additional information necessary for a
morphodynamic run.
Users of Delft3D-FLOW are familiar with two versions of these files: with or without keywords.
D-Flow 1D uses the version with keywords. Besides the <∗.sed> and <∗.mor> file SOBEK 3
might require the following files:
The sediment layer file (<∗.sdb>) contains information about the thickness of a sediment
T
layer.
The sediment diameter file (<∗.d50>) can be used for spatially varying sediment diameters.
DR
AF
The sediment transport and morphology boundary condition file (<∗.bcm>).
The sediment transport file (<∗.tra>).
The nodal relation file (<∗.nrd>) is used to define the function governing the sediment
distribution on nodal points with two or more outflowing branches (bifurcations, trifurcations,...) and any number of inflowing branches.
A table file (<∗.tbl>) can be used for additional control over the sediment distribution at
bifurcations.
The <∗.sed> and <∗.mor> can be generated with Delft3D-FLOW and/or a regular text editor.
The details are described in Appendix section B.1. Before running the model, the files must
be placed in the (<∗.dsproj_data>) directory.
Restrictions:
SOBEK 3 does not yet support fixed layer modelling
SOBEK 3 does not yet support multiple sediment fractions (graded sediment)
Before activating a model run, the files must be placed in the directory: <∗.dsproj_data>.
6.3
Output files
The output file (<morph-gr.his>) will be placed in the (<∗.dsproj_data/water_flow_1d_output>)
directory. This file can be inspected or processed with tools that can handle (<∗.his>) file,
like ODS view or Python Scripting. There are also MatLab functions freely available from the
open repository Open Earth (www.openearth.eu).
6.4
Scripting support
Delta Shell allows the user to extend the functionality of the modelling suite via Python Scripting. This applies to Morphology and Sediment Transport as well. SOBEK 3 comes (since
version 3.3) equipped with several scripts that extend the functionality of Delta Shell and
SOBEK 3. The following paragraphs show several examples how to
100
Deltares
Module D-Flow 1D: Morphology and Sediment Transport
6.4.1
Generating input files and working with spatially varying input
The input files for spatially varying input (<∗.d50> and <∗.sdb>) are generally difficult - if
not impossible - to generate outside DeltaShell. To help with the setup of a morphological
simulation use the <SobekMorphology> class. The following example shows how to quickly
setup morphological files for a fictional model.
from SobekMorphology import MorSetup
# Create a Sobek morphology helper class
SM = MorSetup()
T
#region: Quick setup
# This region shows how to quickly setup spatially varying
# input for morphology. By default the sediment thickness
# is 10 m and the mean sediment diameter is 0.014 m.
# Change the default d50 sediment diameter of branch 'Channel1' to 8 mm.
SM.branch["ChannelName1"].set_uniform_d50(0.08)
6.4.2
DR
AF
# Create input files for morphology
SM.create_input_files()
#endregion
Dumping and dredging
In reality river managers intervene in the natural system in several ways. Dredging — the
removal of sediment from the river bed — is a common channel maintenance intervention.
This might be coupled with subsequent dumping, i.e. the reallocation of the dredged sediment to other parts of the river. Dumping and dredging is not (yet) supported in the computational core, in contrast with Delft3D-FLOW. Alternatively, this functionality is offered via Python
scripting via the <SobekDredgeDump> class. Delta Shell comes with several examples of
how to work with Dredging and Dumping.
The output file (<morph-gr.his>) will be placed in the directory: <∗.dsproj_data/water_flow_1d_output>.
This file can be inspected or processed as any SOBEK 3 history (<∗.his>) file.
Deltares
101
DR
AF
T
SOBEK 3, User Manual
102
Deltares
References
Bailard, J. A., 1981. “An Energetics Total Load Sediment Transport Model for Plane Sloping
Beaches.” Journal of Geophysical Research 86 (C11): 10938-10954.
Becker, B. and Q. Gao, 2012. “Multiple model coupling through OpenMI.” Deltares-memo No.
1205954-003-ZWS-0006.
Becker, B. and G. Prinsen, 2010. “Quasi-(in)stationaire berekeningen met Sobek (steady
simulation mode).” Deltares-memo No. 1202134-011-ZWS-0002. In Dutch.
Deltares, 2012. SOBEK online help. Distributed with SOBEK 2.12.
T
Deltares, 2013. SOBEK 3 / Hydrodynamics Technical Reference Manual / SOBEK in Delta
Shell. Deltares, Delft. Version: 3.0.1.27817.
DR
AF
Gaeuman, D., E. Andrews, A. Krause and W. Smith, 2009. “Predicting fractional bed load
transport rates: Application of the Wilcock-Crowe equations to a regulated gravel bed river.”
Water Resources Research 45.
Grasmeijer, B. and L. Van Rijn, 1998. “Breaker bar formation and migration.” Coastal Engineering pages 2750-2758. Virginia, USA.
Isobe, M. and K. Horikawa, 1982. “Study on water particle velocities of shoaling and breaking
waves.” Coastal Engineering in Japan 25: 109-123.
Nipius, K. G., 1998. Transverse transport modelling using Bailard applied to Grevelingenmouth delta. Delft University of Technology, Delft, The Netherlands. M.Sc. thesis, in
Dutch (Dwarstransportmodellering m.b.v. Bailard toegepast op de Voordelta Grevelingenmonding).
Rienecker, M. M. and J. D. Fenton, 1981. “A Fourier approximation method for steady water
waves.” Journal of Fluid Mechanics 104: 119-137.
Rijn, L. C. van, 1984a. “Sediment transport, Part I: bed load transport.” Journal of Hydraulic
Engineering 110 (10): 1431-1456.
Rijn, L. C. van, 1984b. “Sediment transport, Part II: suspended load transport.” Journal of
Hydraulic Engineering 110 (11): 1613-1640.
Rijn, L. C. van, 1984c. “Sediment transport, Part III: bed form and alluvial roughness.” Journal
of Hydraulic Engineering 110 (12): 1733-1754.
Rijn, L. C. van, 1993. Principles of Sediment Transport in Rivers, Estuaries and Coastal Seas.
Aqua Publications, The Netherlands.
Rijn, L. C. van, 2001. General view on sand transport by currents and waves : data analysis
and engineering modelling for uniform and graded sand (TRANSPOR 2000 and CROSMOR 2000 models). Z2899.20 / Z2099.30 / Z2824.30. WL | Delft Hydraulics, Delft, The
Netherlands.
Rijn, L. C. van, 2003. “Sediment transport by currents and waves; general approximation
formulae Coastal Sediments.” In Corpus Christi, USA.
Rijn, L. C. van, J. A. Roelvink and W. T. Horst, 2000. Approximation formulae for sand transport
by currents and waves and implementation in DELFT-MOR. Tech. Rep. Z3054.40, WL |
Delft Hydraulics, Delft, The Netherlands.
Deltares
103
SOBEK 3, User Manual
Rijn, L. van, D. Walstra, B. Grasmeijer, J. Sutherland, S. Pan and J. Sierra, 2003. “The
predictability of cross-shore bed evolution of sandy beaches at the time scale of storms
and seasons using process-based profile models.” Coastal Engineering 47: 295-327.
RIZA, 2005. Salt Intrusion Technical Reference. RIZA Institute for Inland Water Management
and Waste Water Treatment. Delivered with SOBEK-RE 2.52.005.
Roelvink, J. A. and M. J. F. Stive, 1989. “Bar-generating cross-shore flow mechanisms on a
beach.” Journal of Geophysical Research 94 (C4): 4785-4800.
Soulsby, R., 1997. Dynamics of marine sands, a manual for practical applications. Thomas
Telford, London.
T
Soulsby, R. L., A. G. Davies, J. Fredsøe, D. A. Huntley, I. G. Jonnson, D. Myrhaug, R. R.
Simons, A. Temperville and T. J. Zitman, 1993. “Bed shear stresses due to combined
waves and currents.” In Abstracts-in-depth of the Marine Science and Technology G8-M
overall workshop, Grenoble., pages 2.1-1/2.1-4.
DR
AF
Stelling, G. S. and S. P. A. Duinmeijer, 2003. “A staggered conservative scheme for every Froude number in rapidly varied shallow water flows.” International Journal Numerical
Methods In Fluids 43: 1329-1354.
Stelling, G. S. and A. Verwey, 2006. “Numerical flood simulation.” In Encyclopedia of Hydrological Sciences. John Wiley & Sons.
Stive, M. J. F., 1986. “A model for cross-shore sediment transport.” In Proceedings 20th
International Coastal Engineering Conference, pages 1550-1564. American Society of Civil
Engineers, New York.
Swart, 1974. Offshore sediment transport and equilibrium beach profiles. Ph.D. thesis, Delft
University of Technology, Delft, The Netherlands. Delft Hydraulics Publ. 131.
Thatcher, M. L. and D. R. F. Harleman, 1972. A mathematical model for the prediction of
unsteady salinity intrusion in estuaries. Report no. 144, MIT School of Engineering Massachusetts Institute of Technologie, Department of Civil Engineering.
Wikipedia, 2010.
“Salt water intrusion.”
Saltwater_intrusion.
URL http://en.wikipedia.org/wiki/
Wilcock, P. and J. Crowe, 2003. “Surface-based transport model for mixed-size sediment.”
Journal of Hydraulic Engineering 129 (2): 120-128.
104
Deltares
A How to use OpenDA for Delta Shell models
A.1
Introduction
SOBEK 3 is a modelling system based on the newly developed Delta Shell framework. Modules can be plugged-in. D-Flow 1D and D-RealTimeControl (D-RTC) are the most characteristic modules for SOBEK 3.
The calibration and Ensemble Kalman Filtering (EnKF) of the SOBEK 3 models is done by
means of OpenDA (see www.openda.org). This is a generic functionality and as such part of
Delta Shell. In this document we will speak of Delta Shell (models).
DR
AF
T
Both the calibration of Delta Shell models and running them in EnKF-mode is done by using
OpenDA. To run an OpenDA calibration or EnKF-simulation, a so called OpenDA application
(.oda) file is needed, in which the application to be performed is specified. This oda file is
the top of a hierarchy of configuration files that is organized in a directory structure that is
usually setup as indicated below. The underlined filenames indicate the files that are related
to preparing a Delta Shell model for OpenDA.
topDir (containing e.g. <main_calibration_config.oda>)
algorithm contains the configuration file(s) for the calibration algorithm
stochObserver contains the configuration file(s) and measurement data for the so
called ‘stochastic observer’, the set of measures and the specification of their uncertainty
stochModel contains the configuration file(s) for the so called ‘stochastic model factory’, that specify how model instances can be created. For Delta Shell models, this is
described in
◦ stochModel.xml describes which items can be calibrated, and specifies the re-
lation between the measurement series and the related observation point in the
model
◦ modelConfig.xml specifies the Delta Shell model (the <∗.dsproj>-file and the
name of the model in that project), and some other optional settings for repeatedly
running the model.
For the all over structure and the content of the various files, the user is referred to the documentation of OpenDA on www.openda.org. The two underlined files are described in the
sections below.
A.2
A.2.1
The Stochastic Model configuration
Configuration for calibration
For calibration, the stochastic model configuration file (<stochModel.xml>) specifies which
items can be calibrated, and specifies the relation between the measurement series and the
related observation point in the model. Typically the content of this file looks like the example
below (the grey lines are standard, i.e. they will always be the same):
<?xml version="1.0" encoding="UTF-8"?>
<blackBoxStochModel xmlns:oda="http://www.openda.org"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
Deltares
105
SOBEK 3, User Manual
DR
AF
T
xsi:schemaLocation="http://www.openda.org
http://www.openda.org/schemas/blackBoxStochModelConfig.xsd">
<modelFactory className="org.openda.dotnet.ModelFactoryN2J" workingDirectory=".">
<arg>
OpenDA.DotNet.OpenMI.Bridge.ModelFactory;
DeltaShell.OpenDaOnOpenMI2Wrapper.DeltaShellOpenDAModelProvider
</arg>
</modelFactory>
<vectorSpecification>
<parameters>
<regularisationConstant>
<stdDev value=".1" transformation="ln"/>
<vector id="Kalkmas1_A.x0.q200.Chezy"/>
<vector id="Kalkmas1_A.x3718.q200.Chezy"/>
<vector id="Kalkmas1_B.x0.q200.Chezy"/>
</regularisationConstant>
<regularisationConstant>
<stdDev value=".1" transformation="ln"/>
<vector id="Kalkmas2.x0.q200.Chezy"/>
<vector id="Kalkmas2.x2203.q200.Chezy"/>
<vector id="Grensms1.x0.q200.Chezy"/>
</regularisationConstant>
....
</parameters>
<predictor>
<vector id="H\_Eijsden\_grens.waterlevel"
sourceVectorId="H\_Eijsden\_grens.h"/>
<vector id="H\_Maastricht\_(St.Piet).waterlevel"
sourceVectorId="H\_Maastricht\_(St.Piet).h"/>
</predictor>
</vectorSpecification>
</blackBoxStochModel>
The ‘regularisationConstant’ blocks indicate which roughness sections can be calibrated. The
calibration algorithm treats sections that are grouped in one regularisationConstant block as
one parameter, meaning that they are modified in the same way (i.e. multiplied by the same
factor) .
A.2.2
Configuration for Ensemble Kalman Filtering
For EnKF, the stochastic model configuration file (stochModel.xml) specifies the state of the
model. This state is a combination of a part of the model’s computational state (for SOBEK 3
models, this is the computed water level) and the so called noise models, the models that impose noise on the boundary conditions and/or the state. The second part of the configuration
specifies the relation between the measurement series and the related observation point in
the model. Typically the content of this file looks like the example below (the grey lines are
standard, i.e. they will always be the same):
<?xml version="1.0" encoding="UTF-8"?>
<blackBoxStochModel xmlns="http://www.openda.org"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.openda.org
http://schemas.openda.org/blackBoxStochModelConfig.xsd">
<modelFactory className="org.openda.dotnet.ModelFactoryN2J" workingDirectory=".">
<arg>
DeltaShell.OpenDaWrapper.DeltaShellOpenDAModelFactory;wrapperConfig.xml
</arg>
</modelFactory>
<vectorSpecification>
106
Deltares
How to use OpenDA for Delta Shell models
A.3
DR
AF
T
<state>
<noiseModel id="boundaryNoiseModel"
className="org.openda.noiseModels.TimeSeriesNoiseModelFactory"
workingDirectory=".">
<configFile>boundaryNoise.xml</configFile>
<exchangeItems>
<exchangeItem id="upStreamBoundary.Q" operation="add">
<modelExchangeItem id="QBoundary.Node001.water_discharge"/>
</exchangeItem>
</exchangeItems>
</noiseModel>
<vector id="state" />
</state>
<predictor>
<vector id="ObservationPoint1.waterlevel"
sourceVectorId="ObservationPoint.ObservationPoint1.water_level" />
</predictor>
</vectorSpecification>
</blackBoxStochModel>
The Model configuration
The <modelConfig.xml> file in the stochModel directory specifies which model in which
<∗.dsproj>-file has to be calibrated or to be run in EnKF-mode, and also contains some
additional (often optional) info on how to manage the model computations that are repeatedly
invoked by the algorithm. The table below describes the fields in the xml file. The file looks
like this for calibration:
<?xml version="1.0" encoding="UTF-8"?>
<DeltaShellOpenDAModelProviderSettings
xmlns:xsi=http://www.w3.org/2001/XMLSchema-instance
xmlns:xsd="http://www.w3.org/2001/XMLSchema">
<ProjectPath>
d:\deltaShell\openda\j03\_16138_run_v062.dsproj
</ProjectPath>
<ModelName>
Integrated model placeMaas
</ModelName>
<WorkDirectoryRTC>
.\textbackslash WorkRTC
</WorkDirectoryRTC>
<ModelInstancesCloneDir>
.\textbackslash instances
</ModelInstancesCloneDir>
<KeepEngineDirectories>
true
</KeepEngineDirectories>
</DeltaShellOpenDAModelProviderSettings>
and like this for EnKF:
<?xml version="1.0" encoding="UTF-8"?>
<DeltaShellOpenDAWrapperConfig
xmlns:xsi=http://www.w3.org/2001/XMLSchema-instance
xmlns:xsd="http://www.w3.org/2001/XMLSchema">
<ProjectPath>
d:\deltaShell\openda\j03\16138_run_v062.dsproj
</ProjectPath>
<ModelName>
Integrated model Maas
Deltares
107
SOBEK 3, User Manual
</ModelName>
</DeltaShellOpenDAWrapperConfig>
Table A.1: Description of XML tags
Description
Remarks
ProjectPath
The path of the <∗.dsproj> file, either as full path, or specified relative to the modelConfig.xml file
Must be present
ModelName
The name of the model in the
<∗.dsproj>-file, i.e. the model’s
name in the project explorer
Must be present
ModelInfoForOpenDaFilePath
Output file for providing information
for OpenDA on what items can be
calibrated, and what observations
points are available
Optional
CalibrationValuesLogFilePath
Output file for logging per model
evaluation the actual values of the
calibrated parameters
Optional
EnKFLogFilePath
Log file for Ensemble Kalman Filtering
Prepared
for
logging,
but
no
additional
logging needed
yet (OpenDA’s
main result file
suffices)
RequestedOutPutItemsFile
File specifying which output items
(i.e. quantities at result locations)
should be provided by the main
model (i.e. the ’average’ model of
the filtering process)
EnKF
Optional
only.
RequestedOutPutItemsResultFile
File to which the results mentioned
above should be written.
EnKF
Optional
only.
WorkDirectoryFlow1D
Directory where the D-Flow1D
model engine should store it’s temporary files when running a model
instance. Subdirectory of one of
the model instance directories (see
ModelInstancesCloneDir below)
Optional
but
recommended(to
avoid too many
runs on TEMPdir)
DR
AF
T
Variable
108
Deltares
How to use OpenDA for Delta Shell models
Table A.1: Description of XML tags
Description
Remarks
WorkDirectoryRTC
Directory where the Real Time
Control model engine should store
it’s temporary files when running a
model instance. Subdirectory of
one of the model instance directories (see ModelInstancesCloneDir
below)
Optional
but
recommended
KeepEngineDirectories
If set to true, the engine’s working
directories mentioned above are
not deleted after the run (available
for debugging purposes)
Optional(default
false)
KeepStateFiles
If set to true, the model state files
are not deleted after the run (available for debugging purposes)
EnKF only. Optional
(default false)
Keep1DStateXyzFiles
If set to true, the SOBEK 3-Flow1D
state files are not deleted after the
run (available for debugging purposes)
EnKF only. Optional
(default false)
UseMemoryClone
If set to true, the model is cloned
in memory, instead of repeatedly copying the <∗.dsproj>-file
and loading the model from the
<∗.dsproj>-file
Optional(default
false)
ModelInstancesCloneDir
Directory that serves as a parent directory for the instance directories
that are created for each copy of
the <∗.dsproj>-file (calibration) or
ensemble member (EnKF).
Has to be set
when UseMemoryClone is set
to false
RunnerInstancesCloneDir
Directory that serves as a parent directory for the directories that are
created for running an ensemble
member computations.
EnKF only
NumProcessors
The number of ’runners’ that are
available for running the ensemble
members.
EnKF only. Optional
(default 1)
CleanupInstances
If the model instances are produced by copying the <∗.dsproj>file (i.e. UseMemoryClone is false),
this flag indicates whether these
copied <∗.dsproj>-file’s should be
deleted or not
Optional(default
false)
DR
AF
T
Variable
Deltares
109
SOBEK 3, User Manual
Both directories and files can be specified as either a full path, or as a path relative to the
modelConfig.xml file.
Note: For EnKF the <modelConfig.xml>-file may also be named <wrapperConfig.xml>.
A.4
Installing OpenDA for Delta Shell models
Both the OpenDA calibration application and the OpenDA EnKF application for Delta Shell
models are distributed as part of the SOBEK 3 installation.
A.5
DR
AF
See next Section on how to start the application.
T
Both executables (<DeltaShell.OpenDaCalApplication.exe> and <DeltaShell.OpenDaEnKFApplication.exe>)
are available in the same <bin> directory as where <DeltaShell.Gui.exe> is. Both applications can also be copied out of the zip file <OpenDaApplication.zip>, same <bin> directory
as above.
Running the OpenDA application
To run the application, go to the directory where <DeltaShell.OpenDaCalApplication.exe>
and <DeltaShell.OpenDaEnKFApplication.exe> are, and start DeltaShell.OpenDaCalApplication
or DeltaShell.OpenDaEnKFApplication with only one argument, the full path of the OpenDA
application file (<*.oda>, see section A.1):
> DeltaShell.OpenDaCalApplication.exe ...\myOdaFile.oda
DeltaShell.OpenDaCalApplication also has an option to only extract a list of input parameters
(roughness sections) and output variables (discharge and water level at observation points).
This facilitates setting up the stochModel.xml config file mentioned in section A.2.
To achieve this, start the executable with the following arguments:
DeltaShell.OpenDaApplication projectPath modelName [outFile]
Table A.2: OpenDA program arguments
Argument
Description
Remarks
projectPath
The full path of the <∗.dsproj> file
Must be present
modelName
The name of the model in the <∗.dsproj> file, i.e.
the model’s name in the project explorer
Must be present
outFile
Output file that provides information for OpenDA on
what items can be calibrated, and what observations
points are available.
It this argument is omitted, the file ‘model-info-foropenda.txt’ will be written (in the same directory as
the <∗.dsproj>-file).
Optional
110
Deltares
B Appendix: Morphology and Sediment Transport
Sediment input file
The sediment input file contains the characteristics of all sediment fractions. In the record
description the name of the quantities are given to simplify their reference in the formulas
given in section B.3.
Remark:
Users of Delft3D-FLOW are familiar with two versions of the <∗.sed> file: with or without keywords. D-Flow 1D uses the keyword based version which is described in Table B.1.
T
B.1.1
Input files
Restrictions:
SOBEK 3 does not yet support fixed layer modelling
SOBEK 3 does not yet support multiple sediment fractions (graded sediment)
DR
AF
B.1
Table B.1: Sediment input file with keywords
Keyword
Record description
SedimentFileInformation
FileCreatedBy
contains version number of FLOW-GUI
FileCreationDate
creation date and time of the <∗.sed> file
FileVersion
version number
SedimentOverall
NodeRelations
file specifying node relations <∗.nrd>
Sediment
Name
name between # as specified in NamC in mdf-file
SedTyp
type of sediment; must be “sand” or “bedload”: (1 string)
RhoSol
specific density of sediment fraction [kg/m3 ] (1 real)
SedDxx
xx percentile sediment diameter (for sand or bedload)
where xx can take on values from 01 to 99 [m] (1 real)
SedMinDia
minimum sediment diameter (for sand or bedload) [m]
(1 real)
continued on next page
Deltares
111
SOBEK 3, User Manual
Table B.1 – continued from previous page
Record description
SedDia
median sediment diameter (for sand or bedload) equivalent
to SedD50 [m]
uniform value (1 real)
or file <∗.d50> with spatially varying values at cell centres
(1 string)
SedSg
geometric standard deviation of sediment diameter (for
sand or bedload) [m] (1 real)
SedMaxDia
maximum sediment diameter (for sand or bedload) [m]
(1 real)
T
Keyword
dry bed density [kg/m3 ] (1 real)
CDryB
initial sediment mass at bed per unit area [kg/m2 ]
initial sediment layer thickness at bed [m]
uniform value (1 real)
or filename <∗.sdb> with non-uniform values at cell centres
(1 string)
DR
AF
SdBUni or IniSedThick
Name of fraction specific sediment transport formula (for
sand or bedload)
TraFrm
Table B.2: Options for sediment diameter characteristics
Specified quantities
Assumptions
SedDia (uniform value) or
SedD50 (uniform value)
Piecewise log-uniform distribution
SedD10 = 0.75 SedD50
SedD90 = 1.5 SedD50
SedDia
(filename)
SedD50 (filename)
or
Lognormal distribution (spatially varying grain size)
SedSg = 1.34
SedDia
(filename)
or
SedD50 (filename), SedSg
Lognormal distribution (spatially varying grain size)
SedDxx (any xx), SedSg
Lognormal distribution
Two SedDxx values
Lognormal distribution
SedSg computed from xx and SedDxx
SedMinDia, SedMaxDia
Loguniform distribution
continued on next page
112
Deltares
Appendix: Morphology and Sediment Transport
Table B.2 – continued from previous page
Assumptions
SedMinDia or SedMaxDia,
One SedDxx value
Loguniform distribution
More than two SedDxx,
SedMinDia, SedMaxDia values
Piecewise loguniform distribution
Other combinations
not allowed
Example of a version 2 file, with keywords:
T
Specified quantities
B.1.2
DR
AF
[SedimentFileInformation]
FileCreatedBy
= Delft3D-FLOW-GUI, Version: 3.39.14.03
FileCreationDate = Thu Dec 08 2005, 14:47:46
FileVersion
= 02.00
[SedimentOverall]
IopSus
= 0
Suspended sediment size is Y/N
calculated dependent on d50
Cref
= 1.60e+03
[kg/m3] CSoil Reference density for hindered
settling
[Sediment]
Name
= #Sediment1#
Name of sediment
SedTyp
= bedload
Must be "sand" or "bedload"
RhoSol
= 2.6500000e+003 [kg/m3] Specific density
SedDia
= 2.0000000e-004 [m]
Median sediment diameter (D50)
CDryB
= 1.6000000e+003 [kg/m3] Dry bed density
IniSedThick
= 0.50e+000
[m]
Initial sediment layer thickness at bed
(uniform value or file name)
FacDSS
= 1.0e+0
[-]
FacDss*SedDia = Initial suspended
sediment diameter.
Morphology input file
The morphological input file contains additional information necessary for a morphodynamic
run. Users of Delft3D-FLOW are familiar with two versions of the file, like the <∗.sed> file:
with or without keywords. D-Flow 1D uses the version with keywords.
Table B.3: Morphological input file with keywords
Keyword
Record description
MorphologyFileInformation
FileCreatedBy
contains version number of FLOW-GUI
FileCreationDate
creation date and time of the <∗.mor> file
FileVersion
version number
continued on next page
Deltares
113
SOBEK 3, User Manual
Table B.3 – continued from previous page
Keyword
Record description
Morphology
morphological scale factor
constant (1 real) or
file with time-dependent values (string)
in case of a file: no text may be used after the
filename
MorStt
time interval in minutes after the start of the simulation after which morphological changes will be calculated (1 real)
BedUpd
update bed level during flow run (1 logical: false or
true)
DR
AF
BcFil
T
MorFac
file containing morphological boundary conditions (1
string)
Boundary
Name
IBedCond
name of boundary node ID (1 string)
bedload or bed level boundary condition (1 integer in
the range 0 to 5)
0
no bed level constraint
1
bed level fixed
2
depth specified as function of time
3
depth change specified as function of
time
4
bedload transport rate prescribed (volume rate of bed material)
5
bedload transport rate prescribed (volume rate of stone)
the Boundary block can be repeated for other boundaries
For these boundary conditions you need to specify the imposed time-series in the file referred
to using the BcFil keyword. File format described in section B.1.4.
Example of a version 2 file, with keywords:
[MorphologyFileInformation]
FileCreatedBy
= Delft3D-FLOW-GUI, Version: 3.39.14.03
FileCreationDate = Thu Dec 08 2005, 14:47:50
FileVersion
= 02.00
[Morphology]
MorFac
= 1.0000000e+000 [-] Morphological scale factor
MorStt
= 7.20e+02
[min] Spin-up interval from TStart till start of morph changes
114
Deltares
Appendix: Morphology and Sediment Transport
Table B.4: Additional transport relations
Bedload
Affected
by Waves
IFORM
B.4.1, Van Rijn (1993)
B.4.2, Engelund-Hansen (1967)
B.4.3, Meyer-Peter-Muller (1948)
B.4.4, General formula
B.4.5, Bijker (1971)
B.4.6, Van Rijn (1984)
B.4.7, Soulsby/Van Rijn
B.4.8, Soulsby
B.4.9, Ashida–Michiue (1974)
B.4.10, Wilcock–Crowe (2003)
B.4.11, Gaeuman et al. (2009) laboratory calibration
B.4.12, Gaeuman et al. (2009) Trinity
River calibration
User-defined
Bedload + suspended
Total transport
Total transport
Total transport
Bedload + suspended
Bedload + suspended
Bedload + suspended
Bedload + suspended
Total transport
Bedload
Bedload
Yes
No
No
No
Yes
No
Yes
Yes
Yes
No
No
-1
1
2
4
5
7
11
12
14
16
17
Bedload
No
18
Yes
—
DR
AF
T
Formula
BedUpd
= true
BcFil
= #dmor.bcm#
[Boundary]
Name
= #Node001#
IBedCond = 4
[Boundary]
Name
= #Node002#
IBedCond = 0
Bedload + suspended
Update bathymetry during flow run
Name of morphological boundary condition file
Boundary node ID
0: free none - 1: fixed none - 2: time series depth m
3: depth change prescribed depth change m/s
4: transport incl pores prescribed transport incl pores m3/s
5: transport excl pores prescribed transport excl pores m3/s
Boundary node ID
Remark:
The file for specifying bedload, bed level and/or bed composition boundary conditions
is described in section B.1.4.
Restriction:
The values of the parameters are not checked against their domains.
B.1.3
Sediment transport input file
By default, the formulations of Van Rijn et al. (2000) are applied for the suspended and bedload transport of non-cohesive sediment. In addition this feature offers a number of extra
sediment transport relations for non-cohesive sediment; Table B.4 gives an overview of those
additional formulae.
If you want to use one of these formulae, you must create a sediment transport input file
<∗.tra>. The filename of the sediment input file must be specified in the sediment input
file (formula used for only the selected non-cohesive sediment fraction) using the keyword
TraFrm. In the former case, use the Data Group Addition parameters in the FLOW-GUI
Deltares
115
SOBEK 3, User Manual
with keyword and value: TraFrm=#name.tra#. For these pre-defined alternative transport
relations the sediment transport input file should comply with the following specifications:
The file may start with an arbitrary number of lines not containing the text IFORM.
Then a line starting with sediment transport formula number IFORM and containing text
IFORM.
Then an arbitrary number of lines starting with an asterisk (∗) may follow.
Then a line starting with the number sign (#) followed by a transport formula number
T
optionally followed by text identifying the transport formula for the user. The next lines
should contain the parameter values of the transport formula coefficients: one parameter
value per line optionally followed by text identifying the parameter. There may be an
arbitrary number of blocks starting with # in the file, but exactly one should correspond to
the transport formula number IFORM specified above.
DR
AF
An example file for transport formula 5 referred to as “Bijker (1971)” is provided below. The
following table lists the parameters to be specified in the sediment transport input file for each
separate transport formula.
Table B.5: Transport formula parameters
Formula
Parameter
B.4.1, Van Rijn (1993)
none
B.4.2, Engelund-Hansen (1967)
calibration coefficient α
bed roughness height rk (dummy)
m
B.4.3, Meyer-Peter-Muller (1948)
calibration coefficient α
dummy argument
NA
B.4.4, General formula
calibration coefficient α
power b
power c
ripple factor or efficiency factor µ
critical mobility factor θc
-
calibration coefficient b for shallow water
-
B.4.5, Bijker (1971)
Unit
BS
calibration coefficient b for deep water
-
BD
shallow water (hw /h) criterion Cs
deep water (hw /h) criterion Cd
dummy argument
bed roughness height rc
settling velocity w
porosity ε
wave period Tuser (used if computed
wave period < 10−6 )
NA
m
m/s
s
B.4.6, Van Rijn (1984)
calibration coefficient α1
dummy argument
reference level ξc
settling velocty ws
NA
m
m/s
B.4.7, Soulsby/Van Rijn
calibration coefficient Acal
-
116
Deltares
Appendix: Morphology and Sediment Transport
Unit
D90 /D50 ratio
z0 roughness height
m
B.4.8, Soulsby
calibration coefficient Acal
model index modind
D50 /z0 ratio χ
-
B.4.9, Ashida–Michiue (1974)
calibration coefficient α
critical mobility factor θc
power m
power p
power q
-
B.4.10, Wilcock–Crowe (2003)
none
B.4.11, Gaeuman et al. (2009) laboratory calibration
calibration coefficient θc0
-
calibration coefficient α0
-
T
Parameter
DR
AF
Formula
B.4.12, Gaeuman et al. (2009) Trinity
River calibration
User-defined
-
calibration coefficient θc0
-
calibration coefficient α0
none
-
Remarks:
Van Rijn (1993) does not require any additional parameters. Only the transport formula
number (-1) followed by the string IFORM is required.
The user-defined transport formula requires a keyword based transport input file as
described below.
The keyword IFORM must be present in the same line as the formula number.
The file should not contain tabs.
Example for Engelund-Hansen (1967) formula:
1
IFORM
#1
ENGELUND-HANSEN
1.00
0.0
Example for Meyer-Peter Mueller (1948) formula:
2
IFORM
#2
MEYER-PETER-MULLER
1.00
0.0
Example for Van Rijn 1993 formula:
This is an example of a sediment transport input file to be used with the Van Rijn (1993)
formulations in which case there is no need for additional parameters. The file could simply
read:
Deltares
117
SOBEK 3, User Manual
-1
Sediment transport and morphology boundary condition file
The bcm file contains time-series for sediment transport and morphology boundary conditions.
For each open boundary segment that according to the boundary characteristics given in the
<∗.mor> file requires boundary data, the data is given in two related blocks:
A header block containing a number of compulsory and optional keywords accompanied
Description header block:
Text
Required
Value
table-name
no
arbitrary string
T
by their values.
A data block containing the time dependent data.
DR
AF
B.1.4
IFORM
location
yes
’boundary node ID’
time-function
no
{’non-equidistant’} or ’equidistant’
time-step
yes
time step only in case of time-function ’equidistant’
reference-time
yes
yyyymmdd, yyymmdd hhmmss or ’from model’
time-unit
no
’years’, ’decades’, ’days’, ’hours’, {’minutes’}, ’seconds’, ’ddhhmmss’ or ’date’
interpolation
no
{’linear’} or ’block’
extrapolation
no
’periodic’, ’constant’ or {’none’}
parameter
yes
’parameter name and location’ units ’[ccc]’
records-in-table
no
number of times/lines in the data block
Remarks:
Default parameter values are indicated in braces.
Reference-time not required if time-unit equals ’date’.
Unit strings are currently not interpreted by SOBEK 3.
The ‘parameter name and location’ strings depend on the boundary type chosen, i.e. quantity
type to be specified. The following table lists the base parameter names. The full ‘parameter
name and location’ string is a concatenation of the indicated base parameter name, optionally
followed by a single space character and the user-defined sediment name. Bedload transport should be specified for only the non-mud fractions (i.e. sand and bedload fractions only)
whereas bed composition should be specified for all fractions. The parameters for multiple
sediment fractions must occur in the same order as in which the sediment fractions have
been defined.
118
Deltares
Appendix: Morphology and Sediment Transport
Boundary type
Base
name
parameter
Unit
Multiplicity
m
m/s
m3 /s/m
m3 /s/m
1
1
#nonmud
#nonmud
sediment transport and bed level
0:
1:
2:
3:
4:
5:
free
fixed
time series
depth change prescribed
transport incl pores prescribed
transport excl pores prescribed
none
none
depth
depth change
transport incl pores
transport excl pores
T
Description data block:
Record description
each record
Time in time-units after the reference-time and followed by as many values as parameters have been defined in the description block (all reals).
In case of time-function ‘equidistant’, the first (time) column should be
dropped. In case of time-unit ‘date’ the date and time should be specified as one string using the format: yyyymmddhhmmss.
DR
AF
Record
Remarks:
Maximum record length is 512.
The morphological boundary conditions will only be used at inflow boundaries.
The parameter name of the column should read ‘time’.
Example:
table-name
'Boundary Section : 1'
contents
'Uniform'
location
'Node001'
time-function
'non-equidistant'
reference-time
20141217
time-unit
'minutes'
interpolation
'linear'
parameter
'time' unit '[min]'
parameter
'transport incl pores Sediment1' unit '[m3/s/m]'
records-in-table
2
0.0000
0.000625
6.7108864e+07 0.000625
B.1.5
Nodal Relations Definition file
The nodal relation definition file contains information about the distribution of sediment on
nodal points. A nodal point relation is defined for every node to which three or more branches
are connected, such as bifurcations. By default a proportional function will be used, identical
to Method=function with k=1 and m=0.
Deltares
119
SOBEK 3, User Manual
Table B.6: Nodal relation file with keywords
Keyword
Record description
General
TableFile
name of the tablefile (e.g. ’table.tbl’)
NodalPointRelation
Method
table or function
Table
If method is table, define the name of the table in the
TableFile to use
k
If method is function, the value of the ’k’ parameter
m
If method is function, the value of the ’m’ parameter
BranchIn
The name of the incoming branch
T
name of the node
DR
AF
B.1.6
Node
BranchOut1
Only necessary of method is table. Name of outcoming
branch nr 1
BranchOut2
Only necessary of method is table. Name of outcoming
branch nr 2
Table file
The table file is used to define tables for the nodal relation method ’table’. The file format is
akin to Delft3D-FLOW <∗.pol> or <∗.ldb> files. A table is defined by a name. Each table
consists of two columns and any number of rows. Comments can be inserted by prefixing a
line with an asterix (∗)
The first column of a table file is always defined as the ratio of flow distribution between
BranchOut1 and BranchOut2. The second column is always defined as the ratio of the
sediment distribution. The user should specify in the Nodal Relation File (see B.1.5) which
branch is ’BranchOut1’ and which branch is ’BranchOut2’.
Remark:
The table method can not be used for trifurcations or other situations with more than 2
outflowing branches.
Example table file
* Bifurcation relationship
* column 1 = QBranch1/QBranch2
* column 2 = SBranch1/SBranch2
TABL3
4 2
1.0 1.0
2.0 2.0
120
Deltares
Appendix: Morphology and Sediment Transport
B.2
Output files
T
3.0 2.0
4.0 2.0
* column 1 = QBranch4/QBranch5
* column 2 = SBranch4/SBranch5
TABL6
4 2
1.0 1.0
2.0 2.0
3.0 2.0
4.0 2.0
DR
AF
Morphology and Sediment Transport output is written to <morph-gr.his> files, which are located in the dsproj_data/water_flow_1d_output/work directory.
Delta Shell offers no tool to read <∗.his> files. Users familiar with SOBEK 2 can use ODS
View. The free Open Earth repository has a Matlab scripts available to read <∗.his> files.
B.3
Bedload sediment transport of non-cohesive sediment
Bedload (or, for the simpler transport formulae, total load) transport is calculated for all “sand”
and “bedload” sediment fractions by broadly according to the following approach: first, the
magnitude and direction of the bedload transport at the cell centres is computed using the
transport formula selected (See section B.4), subsequently the transport rates at the cell
interfaces are determined.
B.3.1
Basic formulation
For simulations including waves the magnitude and direction of the bedload transport on a
horizontal bed are calculated using the transport formula selected assuming sufficient sediment and ignoring bed composition except for e.g. hiding and exposure effects on the critical
shear stresses.
The default sediment transport formula is Van Rijn (1993, cf. B.4.1).
Some of the sediment transport formulae prescribe the bedload transport direction whereas
others predict just the magnitude of the sediment transport. In the latter case the initial transport direction will be assumed to be equal to the direction of the characteristic (near-bed) flow
direction.
Deltares
121
SOBEK 3, User Manual
B.3.2
Calculation of bedload transport at open boundaries
B.4
Transport formulations for non-cohesive sediment
T
At open boundaries the user may either prescribe the bed level development or the bedload
transport rates. In the latter case the bedload transport rates are known from the model
input, whereas in the former case the effective bedload transport rates at the boundary could
be derived from the mass balance at the open boundary point. The bed level boundary
condition is imposed at the same location where a water level boundary condition is imposed,
that is at the grid cell just outside the model domain. A consequence of this approach is
that the bed level at the first grid cell inside the model domain will not exactly behave as
you imposed, but in general it will follow the imposed behaviour closely. In case of multiple
sediment fractions, a boundary condition for the bed composition is also needed at inflow
boundaries. See Appendices B.1.2 and B.1.4 for imposing various morphological boundary
conditions.
DR
AF
This special feature offers a number of standard sediment transport formulations for noncohesive sediment. Table B.7 gives a summary of the additional formulae.
Table B.7: Additional transport relations
B.4.1
Formula
Bedload
Waves
B.4.1, Van Rijn (1993)
B.4.2, Engelund-Hansen (1967)
B.4.3, Meyer-Peter-Muller (1948)
B.4.4, General formula
B.4.5, Bijker (1971)
B.4.6, Van Rijn (1984)
B.4.7, Soulsby/Van Rijn
B.4.8, Soulsby
B.4.9, Ashida–Michiue (1974)
B.4.10, Wilcock–Crowe (2003)
B.4.11, Gaeuman et al. (2009) laboratory calibration
B.4.12, Gaeuman et al. (2009) Trinity River calibration
Bedload + suspended
Total transport
Total transport
Total transport
Bedload + suspended
Bedload + suspended
Bedload + suspended
Bedload + suspended
Total transport
Bedload
Bedload
Bedload
Yes
No
No
No
Yes
No
Yes
Yes
No
No
No
No
Van Rijn (1993)
Van Rijn (1993) distinguishes between sediment transport below the reference height a which
is treated as bedload transport and that above the reference height which is treated as
suspended-load. Sediment is entrained in the water column by imposing a reference concentration at the reference height.
Reference concentration
The reference concentration is calculated in accordance with Van Rijn et al. (2000) as:
c(`)
a
122
(`)
(`) 1.5
D50 Ta
= 0.015ρ(`)
s
(`) 0.3
a D∗
(B.1)
Deltares
Appendix: Morphology and Sediment Transport
where:
(`)
ca
mass concentration at reference height a
In order to evaluate this expression the following quantities must be calculated:
non-dimensional particle diameter:
"
(`)
D∗
(`)
Ta
=
(`)
D50
(s(`) − 1)g
ν2
(`)
(`)
(`)
(`)
(µc τb,cw + µw τb,w ) − τcr
(`)
τcr
efficiency factor current:
f 0 (`)
c
fc
DR
AF
µ(`)
c =
f 0 (`)
c
"
(`)
= 0.24
10
12h
log
!#−2
(`)
(`)
10
log
12h
ks
−2
1
= max 0.063,
8
!
Hs 2
1.5 −
h
(B.8)
bed shear stress due to waves:
(B.9)
total wave-related friction factor (≡ Equations ??, B.49 and B.90):

fw = exp −6 + 5.2
Deltares
(B.7)
efficiency factor waves:
2
1
bδ
τb,w = ρw fw U
4
fw
(B.6)
bed shear stress due to current in the presence of waves. Note that the bed
shear velocity u∗ is calculated in such a way that Van Rijn’s wave-current interaction factor αcw is not required.
µ(`)
w
τb,w
(B.5)
total current-related friction factor:
τb,cw = ρw u2∗
µw
(B.4)
3D90
fc(`) = 0.24
τb,cw
(B.3)
gain related friction factor:
(`)
f 0c
fc
(B.2)
non-dimensional bed-shear stress:
Ta(`) =
µc
#1/3
T
(`)
D∗
Aˆδ
ks,w
!−0.19 

(B.10)
123
SOBEK 3, User Manual
To avoid the need for excessive user input, the wave related roughness ks,w is related to the
estimated ripple height, using the relationship:
ks,w = RWAVE · ∆r , with∆r = 0.025 and 0.01 m ≤ ks,w ≤ 0.1 m
(B.11)
where:
RWAVE
(`)
τcr
the user-defined wave roughness adjustment factor. Recommended to be in
the range 1–3, default = 2.
critical bed shear stress:
(`)
(`)
(`)
τcr
= (ρ(`)
s − ρw )gD50 θcr
(`)
(`)
θcr
a
Aˆδ
(`)
D50
(`)
D90
h
ka
ks
ks,w
uz
bδ
U
zu
∆r
δm
δw

0.24D∗−1 ,




 0.14D∗−0.64 ,
0.04D∗−0.1 ,
=


0.013D∗0.29 ,



0.055,
T
threshold parameter θcr is calculated according to the classical Shields curve
as modelled by Van Rijn (1993) as a function of the non-dimensional grain size
D∗ . This avoids the need for iteration.
(`)
Note: for clarity, in this expression the symbol D∗ has been used where D∗
would be more correct:
1 < D∗ ≤ 4
4 < D∗ ≤ 10
10 < D∗ ≤ 20
20 < D∗ ≤ 150
150 < D∗
DR
AF
(`)
θcr
(B.12)
(B.13)
Van Rijn’s reference height
ˆδ =
peak orbital excursion at the bed: A
ˆδ
Tp U
2π .
median sediment diameter
(`)
(`)
90 % sediment passing size: D90 = 1.5D50
water depth
apparent bed roughness felt by the flow when waves are present.
user-defined current-related effective roughness height (space varying)
wave-related roughness, calculated from ripple height, see Equation B.11
velocity magnitude taken from a near-bed computational layer. In a current-only
situation the velocity in the bottom computational layer is used. Otherwise, if
waves are active, the velocity is taken from the layer closest to the height of the
top of the wave mixing layer δ . √
peak orbital velocity at the bed: 2 × RMS orbital velocity at bed, taken from the
wave module.
height above bed of the near-bed velocity (uz ) used in the calculation of bottom
shear stress due to current
estimated ripple height, see Equation B.11
thickness of wave boundary mixing layer following Van Rijn (1993): 3δw (and
δm ≥ ka )
wave boundary layer thickness:
δw = 0.072Aˆδ
ˆδ
A
ks,w
−0.25
.
We emphasise the following points regarding this implementation:
The bottom shear stress due to currents is based on a near-bed velocity taken from the
hydrodynamic calculations, rather than the depth-averaged velocity used by Van Rijn.
All sediment calculations are based on hydrodynamic calculations from the previous half
time-step. We find that this is necessary to prevent unstable oscillations developing.
124
Deltares
Appendix: Morphology and Sediment Transport
The apparent roughness felt by the flow (ka ) is dependent on the hydrodynamic wave-current
interaction model applied. At this time, Van Rijn’s wave-current interaction model is not available in Delft3D-FLOW. This means that it is not possible for a user to exactly reproduce results
obtained using Van Rijn’s full formulations for waves and currents.
Adjustment of the representative diameter of suspended sediment
(`)
The representative diameter of the suspended sediment Ds generally given by the userdefined sediment diameter SEDDIA (D50 of bed material) multiplied by the user-defined factor
FACDSS (see also remarks) can be overruled in case the Van Rijn (1993) transport formula is
selected. This achieved by setting IOPSUS=1 the representative diameter of the suspended
sediment will then be set to:
DR
AF
(`)
T
Ds(`)

(`)
(`)


 0.64D
50
for TA ≤ 1
(`)
(`)
(`)
=
D50 1 + 0.015 TA − 25
for 1 < TA ≤ 25


(`)
 D(`)
for 25 < TA
50
where Ta
(B.14)
is given by equation B.3.
Bedload transport rate
For simulations including waves the magnitude and direction of the bedload transport on a
horizontal bed are calculated using an approximation method developed by Van Rijn et al.
(2003). The method computes the magnitude of the bedload transport as:
(`)
|Sb | = 0.006ρs ws D50 M 0.5 Me0.7
(B.15)
where:
Sb
M
Me
bedload transport [kg m-1 s-1 ]
sediment mobility number due to waves and currents [-]
excess sediment mobility number [-]
M=
2
veff
(s − 1) gD50
(veff − vcr )2
(s − 1) gD50
q
2 + U2
= vR
on
(B.16)
Me =
(B.17)
veff
(B.18)
in which:
vcr
vR
Uon
critical depth averaged velocity for initiation of motion (based on a parameterisation of the Shields curve) [m/s]
magnitude of an equivalent depth-averaged velocity computed from the velocity
in the bottom computational layer, assuming a logarithmic velocity profile [m/s]
near-bed peak orbital velocity [m/s] in onshore direction (in the direction on
wave propagation) based on the significant wave height
Uon (and Uof f used below) are the high frequency near-bed orbital velocities due to short
waves and are computed using a modification of the method of Isobe and Horikawa (1982).
Deltares
125
SOBEK 3, User Manual
This method is a parameterisation of fifth-order Stokes wave theory and third-order cnoidal
wave theory which can be used over a wide range of wave conditions and takes into account the non-linear effects that occur as waves propagate in shallow water (Grasmeijer and
Van Rijn, 1998).
The direction of the bedload transport vector is determined by assuming that it is composed
of two parts: part due to current (Sb,c ) which acts in the direction of the near-bed current, and
part due to waves (Sb,w ) which acts in the direction of wave propagation. These components
are determined as follows:
Sb
Sb,c = p
1 + r2 + 2 |r| cos ϕ
where:
(|Uon | − vcr )3
(|vR | − vcr )3
DR
AF
r=
T
|Sb,w | = r |Sb,c |
(B.19)
(B.20)
(B.21)
Sb,w = 0 if r < 0.01, Sb,c = 0 if r > 100, and ϕ = angle between current and wave direction
for which Van Rijn (2003) suggests a constant value of 90◦ .
Also included in the “bedload” transport vector is an estimation of the suspended sediment
transport due to wave asymmetry effects. This is intended to model the effect of asymmetric
wave orbital velocities on the transport of suspended material within about 0.5 m of the bed
(the bulk of the suspended transport affected by high frequency wave oscillations).
This wave-related suspended sediment transport is again modelled using an approximation
method proposed by Van Rijn (2001):
Ss,w = fSUSW γUA LT
where:
Ss,w
fSUSW
γ
UA
LT
(B.22)
wave-related suspended transport [kg/(ms)]
user-defined tuning parameter
phase lag coefficient (= 0.2)
velocity asymmetry value [m/s] =
4 −U 4
Uon
of f
3
3
Uon +Uof
f
suspended sediment load [kg/m2 ] = 0.007ρs D50 Me
The three separate transport modes are imposed separately. The direction of the bedload
due to currents Sb,c is assumed to be equal to the direction of the current, whereas the two
wave related transport components Sb,w and Ss,w take on the wave propagation direction.
This results in the following transport components:
126
ub,u
|Sb,c |
|ub |
ub,v
=
|Sb,c |
|ub |
Sbc,u =
(B.23)
Sbc,v
(B.24)
Sbw,u = Sb,w cos φ
(B.25)
Sbw,v = Sb,w sin φ
(B.26)
Deltares
Appendix: Morphology and Sediment Transport
Ssw,u = Ss,w cos φ
(B.27)
Ssw,v = Ss,w sin φ
(B.28)
where φ is the local angle between the direction of wave propagation and the computational
grid. The different transport components can be calibrated independently by using the Bed,
BedW and SusW keywords in the morphology input file.
B.4.2
Engelund-Hansen (1967)
S = Sb + Ss,eq = √
0.05αq 5
gC 3 ∆2 D50
where:
magnitude of flow velocity
the relative density (ρs − ρw )/ρw
Chézy friction coefficient
calibration coefficient (O(1))
DR
AF
q
∆
C
α
T
The Engelund-Hansen sediment transport relation has frequently been used in rivers and
estuaries. It reads:
(B.29)
The transport rate is imposed as bedload transport due to currents Sbc . The following formula
specific parameters have to be specified in the input files of the Transport module (see Section
B.1.3): calibration coefficient α and roughness height rk .
Remarks:
The D50 grain size diameter is based on the sediment fraction considered.
A second formula specific input parameter (rk ) is required for the Engelund-Hansen
formula. This parameter, which represents the roughness height for currents alone in
[m], is only used to determine the C value when the Chézy friction in the flow has not
been defined. Generally, this parameter can thus be treated as a dummy parameter.
B.4.3
Meyer-Peter-Muller (1948)
The Meyer-Peter-Muller sediment transport relation is slightly more advanced than the EngelundHansen formula, as it includes a critical shear stress for transport. It reads:
S = 8αD50
p
∆gD50 (µθ − ξθcr )3/2
(B.30)
where:
α
∆
µ
θcr
ξ
calibration coefficient (O(1))
the relative density (ρs − ρw )/ρw
ripple factor or efficiency factor
critical mobility parameter (= 0.047)
hiding and exposure factor for the sediment fraction considered
and the Shields mobility parameter θ given by
θ=
Deltares
q 2
C
1
∆D50
(B.31)
127
SOBEK 3, User Manual
in which q is the magnitude of the flow velocity [m/s]. The ripple factor µ reads:
µ = min
C
Cg,90
1.5
!
, 1.0
(B.32)
where Cg,90 is the Chézy coefficient related to grains, given by:
Cg,90 = 18
10
log
12(d + ζ)
D90
(B.33)
T
with D90 specified in [m]. The transport rate is imposed as bedload transport due to currents
Sbc . The following formula specific parameters have to be specified in the input files of the
Transport module (see Section B.1.3): calibration coefficient α and a dummy value.
B.4.4
DR
AF
Remark:
The D50 is based on the sediment fraction considered, the D90 grain size diameters is
based on the composition of the local sediment mixture.
General formula
The general sediment transport relation has the structure of the Meyer-Peter-Muller formula,
but all coefficients and powers can be adjusted to fit your requirements. This formula is aimed
at experienced users that want to investigate certain parameters settings. In general this
formula should not be used. It reads:
S = αD50
p
∆gD50 θb (µθ − ξθcr )c
(B.34)
where ξ is the hiding and exposure factor for the sediment fraction considered and
θ=
q 2
C
1
∆D50
(B.35)
in which q is the magnitude of the flow velocity.
The transport rate is imposed as bedload transport due to currents Sbc . The following parameters have to be specified in the input files of the Transport module (see Section B.1.3):
calibration coefficient α, powers b and c, ripple factor or efficiency factor µ, critical mobility
parameter θcr .
B.4.5
Bijker (1971)
The Bijker formula sediment transport relation is a popular formula which is often used in
coastal areas. It is robust and generally produces sediment transport of the right order of
magnitude under the combined action of currents and waves. Bedload and suspended load
are treated separately. The near-bed sediment transport (Sb ) and the suspended sediment
transport (Ss ) are given by the formulations in the first sub-section. It is possible to include
sediment transport in the wave direction due to wave asymmetry and bed slope following
the Bailard approach, see Bailard (1981), Stive (1986). Separate expressions for the wave
asymmetry and bed slope components are included:
~b = S
~b0 + S
~b,asymm + S
~s,asymm + S
~b,slope + S
~s,slope
S
~s = S
~s0
S
128
(B.36)
(B.37)
Deltares
Appendix: Morphology and Sediment Transport
where Sb0 and Ss0 are the sediment transport in flow direction as computed according to the
formulations of Bijker in the first sub-section, and the asymmetry and bed slope components
for bedload and suspended transport are defined in the second sub-section. Both bedload and
suspended load terms are incorporated in the bedload transport for further processing. The
transport vectors are imposed as bedload transport vector due to currents Sbc and suspended
load transport magnitude Ss , from which the equilibrium concentration is derived, respectively.
Basic formulation
The basic formulation of the sediment transport formula according to Bijker is given by:
q√
g (1 − φ) exp (Ar )
C
33.0h
+ I2
Ss = 1.83Sb I1 ln
rc
Sb = bD50
T
(B.38)
where
C
h
q
φ
and
(B.39)
Chézy coefficient (as specified in input of Delft3D-FLOW module)
water depth
flow velocity magnitude
porosity
DR
AF
B.4.5.1
Ar = max (−50, min (100, Ara ))
(B.40)
(hw /h) − Cd
b = BD + max 0, min 1,
(BS − BD)
Cs − Cd
(B.41)
I1 = 0.216
I2 = 0.216
rc z∗ −1
h
z
1 − rhc ∗
rc z∗ −1
h
z
1 − rhc ∗
Z1 1−y
y
z∗
dy
(B.42)
rc /h
Z1
ln y
1−y
y
z∗
dy
(B.43)
rc /h
where
BS
BD
Cs
Cd
rc
Coefficient b for shallow water (default value 5)
Coefficient b for deep water (default value 2)
Shallow water criterion (Hs /h) (default value 0.05)
Deep water criterion (default value 0.4)
Roughness height for currents [m]
and
Ara =
Deltares
−0.27∆D50 C 2
2 Ub
2
µq 1 + 0.5 ψ q
(B.44)
129
SOBEK 3, User Manual
µ=
C
10
18 log(12h/D90 )
(B.45)
w
z∗ =
√
κq g
C
r
1 + 0.5
ψ Uqb
(B.46)
2
ωhw
2 sinh (kw h)
(B.47)
2π
T
DR
AF
5.123
fw = exp −5.977 + 0.194
a0
T
Ub =
ω=
1.5
(B.48)
(B.49)
(≡ Equations ??, B.10 and B.90):
Ub
a0 = max 2,
ωrc
(
ψ=
where
C
hw
kw
T
Ub
w
∆
κ
C
0
q
fw
2g
if wave effects are included (T > 0)
(B.50)
(B.51)
otherwise
Chézy coefficient (as specified in input of Delft3D-FLOW module)
wave height (Hrms )
wave number
wave period computed by the waves model or specified by you as T user.
wave velocity
sediment fall velocity [m/s]
relative density (ρs − ρw )/ρw
Von Kármán constant (0.41)
The following formula specific parameters have to be specified in the input files of the Transport module (see Section B.1.3): BS , BD , Cs , Cd , dummy argument, rc , w, ε and T user.
B.4.5.2
Transport in wave propagation direction (Bailard-approach)
If the Bijker formula is selected it is possible to include sediment transport in the wave direction
due to wave asymmetry following the Bailard approach, see Bailard (1981) and Stive (1986).
For a detailed description of the implementation you are referred to Nipius (1998).
Separate expressions for the wave asymmetry and bed slope components are included for
both bedload and suspended load. Both extra bedload and suspended load transport vectors
are added to the bedload transport as computed in the previous sub-section:
~b = S
~b0 + S
~b,asymm + S
~s,asymm + S
~b,slope + S
~s,slope
S
130
(B.52)
Deltares
Appendix: Morphology and Sediment Transport
where the asymmetry components for respectively the bedload and suspended transport in
wave direction are written as:
Sb;asymm (t) =
ρcf εb
|u(t)|2 u(t)
(ρs − ρ) g (1 − φ) tan ϕ
(B.53)
Ss;asymm (t) =
ρcf εs
|u(t)|3 u(t)
(ρs − ρ) g (1 − φ) w
(B.54)
from which the components in ξ and η direction are obtained by multiplying with the cosine
and sine of the wave angle θ w and the bed slope components as:
ρcf εb
1
∂zb
|u(t)|3
(ρs − ρ) g (1 − φ) tan ϕ tan ϕ
∂ξ
(B.55)
Ss;slope,ξ (t) =
ρcf εs
εs
∂zb
|u(t)|5
(ρs − ρ) g (1 − φ) w w
∂ξ
(B.56)
DR
AF
T
Sb;slope,ξ (t) =
and similar for the η direction, where:
u(t)
ρ
ρs
cf
φ
ϕ
w
εb
εs
near bed velocity signal [m/s]
density of water [kg/m3 ]
density of the sediment [kg/m3 ]
coefficient of the bottom shear stress [-] (constant value of 0.005)
porosity [-] (constant value of 0.4)
natural angle of repose [-] (constant value of tan ϕ = 0.63)
sediment fall velocity [m/s]
efficiency factor of bedload transport [-] (constant value of 0.10)
efficiency factor of suspended transport [-] (constant value of 0.02, but in implemented expression for suspended bed slope transport the second εs is replaced
by a user-defined calibration factor; see Equation B.59).
These transports are determined by generating velocity signals of the orbital velocities near
the bed by using the Rienecker and Fenton (1981) method, see also Roelvink and Stive
(1989).
The (short wave) averaged sediment transport due to wave asymmetry, Equations B.53 and
B.54, is determined by using the following averaging expressions of the near bed velocity
signal (calibration coefficients included):
D
E
D
E
D
E
u |u|2 = F acA u
˜ |˜
u|2 + 3F acU u
¯ |˜
u|2
(B.57)
D
E
D
E
D
E
u |u|3 = F acA u
˜ |˜
u|3 + 4F acU u
¯ |˜
u|3
(B.58)
in which:
u
˜
u
¯
F acA
F acU
Deltares
orbital velocity signal
averaged flow velocity (due to tide, undertow, wind, etc.)
user-defined calibration coefficient for the wave asymmetry
user-defined calibration coefficient for the averaged flow
131
SOBEK 3, User Manual
The suspended transport relation due to the bed slope according to Equation B.56 is implemented as:
Ss;slope,ξ (t) =
ρcf εs
εsl
∂zb
|u(t)|5
(ρs − ρ) g (1 − φ) w w
∂ξ
(B.59)
where:
user-defined calibration coefficient EpsSL
εsl
T
To activate this transport option, you have to create a separate file named <coef.inp> which
contains on three separate lines the calibration coefficients: FacA, FacU and EpsSL. The
other parameters are read from the transport input file or are specified as general sediment
characteristics.
Note: the user-defined FacU value is currently treated as a dummy value, FacU = 0.0 will
always be used.
DR
AF
A validation study (Nipius, 1998) showed that the following coefficient settings yielded the best
results for the Dutch coast:
FacA = 0.4
FacU = 0.0
EpsSL = 0.11
If a relatively straight coast is considered the effect of the parameters is:
The wave asymmetry causes onshore directed sediment transport (i.e. in the wave propagation direction). An increased FacA results in an increased onshore transport and hence
steepening of the cross-shore bottom profile.
The bed slope transport is in general offshore directed. By increasing EpsSL an increased
flattening of the bottom profile occurs (i.e. increased offshore transports).
The ratio between these parameters determines the balance between onshore and offshore transport and hence the shape and slope of the cross-shore bottom profile. The
associated response time of the cross-shore morphology can be influenced by modifying
the values of the two parameters, but maintaining a constant ratio. Increased values result
in increased gross transports and consequently a reduced morphological response time
(and vice versa).
B.4.6
Van Rijn (1984)
The Van Rijn (1984a,b,c) sediment transport relation is a transport formula commonly used for
fine sediments in situations without waves. Separate expressions for bedload and suspended
load are given. The bedload transport rate is given by:

q
 0.053 ∆gD3 D−0.3 T 2.1 for T < 3.0
50 ∗
q
Sb =
 0.1 ∆gD3 D−0.3 T 1.5
for T ≥ 3.0
50 ∗
(B.60)
where T is a dimensionless bed shear parameter, written as:
T =
132
µc τbc − τbcr
τbcr
(B.61)
Deltares
Appendix: Morphology and Sediment Transport
It is normalised with the critical bed shear stress according to Shields (τbcr ), the term µc τbc is
the effective shear stress. The formulas of the shear stresses are
1
τbc = ρw fcb q 2
8
0.24
fcb =
10
( log (12h/ξc ))2
10
2
18 log (12h/ξc )
µc =
Cg,90
(B.62)
(B.63)
(B.64)
where Cg,90 is the grain related Chézy coefficient
log
12h
3D90
(B.65)
T
Cg,90 = 18
10
The critical shear stress is written according to Shields:
τbcr = ρw ∆gD50 θcr
(B.66)
DR
AF
in which θcr is the Shields parameter which is a function of the dimensionless particle parameter D∗ :
D∗ = D50
∆g
ν2
1
3
(B.67)
The suspended transport formulation reads:
Ss = fcs qhCa
(B.68)
In which Ca is the reference concentration, q depth averaged velocity, h the water depth and
fcs is a shape factor of which only an approximate solution exists:
fcs =
f0 (zc ) if zc 6= 1.2
f1 (zc ) if zc = 1.2
(ξc /h)zc − (ξc /h)1.2
f0 (zc ) =
(1 − ξc /h)zc (1.2 − zc )
f1 (zc ) =
ξc /h
1 − ξc /h
(B.69)
(B.70)
1.2
ln (ξc /h)
(B.71)
where ξc is the reference level or roughness height (can be interpreted as the bedload layer
thickness) and zc the suspension number:
ws
zc = min 20,
+φ
βκu∗
r
fcb
u∗ = q
8
2 !
ws
β = min 1.5, 1 + 2
u∗
0.8 ws
Ca 0.4
φ = 2.5
u∗
0.65
Deltares
(B.72)
(B.73)
(B.74)
(B.75)
133
SOBEK 3, User Manual
The reference concentration is written as:
Ca = 0.015α1
D50 T 1.5
ξc D∗0.3
(B.76)
The bedload transport rate is imposed as bedload transport due to currents Sbc ,while the
computed suspended load transport rate is converted into a reference concentration equal to
fcs Ca . The following formula specific parameters have to be specified in the input files of the
Transport module (see Section B.1.3): calibration coefficient α1 , dummy argument, reference
level (bedload layer thickness) or roughness height ξc [m] and settling velocity ws [m/s].
T
Soulsby/Van Rijn
The sediment transport relation has been implemented based on the formulations provided in
Soulsby (1997). References in the following text refer to this book.
If the wave period Tp is smaller than 10−6 s, the wave period Tp is set to 5 s and the rootmean-square wave height is set to 1 cm. Furthermore, the wave period is limited to values
larger than 1 s. The root-mean-square wave height is limited to values smaller than 0.4 H ,
where H is the water depth.
DR
AF
B.4.7
The sediment transport is set to zero in case of velocities smaller than 10−6 m/s, water depth
larger than 200 m or smaller than 1 cm.
The root-mean-square orbital velocity is computed as:
Urms =
√
2
πHrms
Tp sinh (kH)
(B.77)
Furthermore, D∗ is defined as (Soulsby, 1997, p.104):
D∗ =
g∆
ν2
1/3
D50
(B.78)
Using the critical bed shear velocity according to Van Rijn (Soulsby, 1997, p.176):
Ucr
(
0.1 10 log (4H/D ) if D
0.19D50
90
50 ≤ 0.5 mm
=
0.6
10
8.5D50 log (4H/D90 ) if 0.5 mm < D50 ≤ 2 mm
(B.79)
larger values of D50 lead to an error and to the halting of the program.
The sediment transport is split into a bedload and suspended load fraction. The direction of
the bedload transport is assumed to be equal to the direction of the depth-averaged velocity
in a 2D simulation and equal to the direction of the velocity at the reference height a (see ??)
in a 3D simulation (Soulsby, 1997, p.183):
134
Sbx = Acal Asb uξ
(B.80)
Sby = Acal Asb vξ
(B.81)
Deltares
Appendix: Morphology and Sediment Transport
and the suspended transport magnitude is given by the following formula (this quantity is
lateron converted to a reference concentration to feed the advection-diffusion equation for the
suspended sediment transport as indicated in ??):
Ss = Acal Ass ξ
p
u2 + v 2
(B.82)
where
a user-defined calibration factor
bedload multiplication factor
Asb = 0.005H
Ass
1.2
suspended load multiplication factor
Ass = 0.012D50
D∗−0.6
(∆gD50 )1.2
a general multiplication factor
(B.83)
DR
AF
ξ
D50 /H
∆gD50
T
Acal
Asb
r
2.4
0.018 2
2
ξ=
U +
U
− Ucr
CD rms
(B.84)
(B.85)
where U is the total depth-averaged velocity and CD is the drag coefficient due
to currents, defined by:
CD =
κ
ln (H/z0 ) − 1
2
(B.86)
where z0 equals 6 mm and the Von Kármán constant κ is set to 0.4.
The bedslope correction factor is not explicitly included in this formula as it is a standard
correction factor available in the online morphology module. The method is intended for conditions in which the bed is rippled.
The following formula specific parameters have to be specified in the input files of the Transport module (See Section B.1.3): the calibration factor Acal , the ratio of the two characteristic
grain sizes D90 /D50 and the z0 roughness height.
B.4.8
Soulsby
The sediment transport relation has been implemented based on the formulations provided in
Soulsby (1997). References in the following text refer to this book.
If the wave period Tp is smaller than 10−6 s, the wave period Tp is set to 5 s and the rootmean-square wave height is set to 1 cm. Furthermore, the wave period is limited to values
larger than 1 s. The root-mean-square wave height is limited to values smaller than 0.4 H ,
where H is the water depth.
The sediment transport is set to zero in case of velocities smaller than 10−6 m/s, water depth
larger than 200 m or smaller than 1 cm.
Deltares
135
SOBEK 3, User Manual
The root-mean-square orbital velocity Urms and the orbital velocity Uorb are computed as
Urms =
√
2Uorb =
√
2
πHrms
Tp sinh (kH)
(B.87)
For a flat, non-rippled bed of sand the z0 roughness length is related to the grain size as
(Soulsby, 1997, eq.25, p.48) where χ is a user-defined constant:
z0 =
D50
χ
(B.88)
a∗ =
Uorb Tp
z0
T
The relative roughness is characterised using a∗ :
(B.89)
which is subsequently used to determine the friction factor of the rough bed according to
Swart (1974) (≡ Equations ??, B.10 and B.49):
0.3
if a∗ ≤ 30π 2
0.00251 exp 14.1a−0.19
if a∗ > 30π 2
∗
DR
AF
fw =
(B.90)
which corresponds to formulae 60a/b of Soulsby (p.77) using r = a∗ /(60π) where r is the
relative roughness used by Soulsby. The friction factor is used to compute the amplitude of
the bed shear-stress due to waves as:
2
τw = 0.5ρfw Uorb
(B.91)
Furthermore, the shear stress due to currents is computed as:
τc = ρCD U 2
where
CD =
κ
1 + ln (z0 /H)
(B.92)
2
(B.93)
as defined on Soulsby (1997, p.53–55). The interaction of the currents and waves is taken
into account using the factor Y in the following formula for mean bed shear stress during a
wave cycle under combined waves and currents (Soulsby, 1997, p.94):
τm = Y (τw + τc )
(B.94)
The formula for Y is given by:
Y = X [1 + bX p (1 − X)q ]
(B.95)
where:
X=
τc
τc + τw
(B.96)
and b is computed using:
b = b1 + b2 |cos φ|J + b3 + b4 |cos φ|J
136
10
log (fw /CD )
(B.97)
Deltares
Appendix: Morphology and Sediment Transport
Table B.8: Overview of the coefficients used in the various regression models (Soulsby
et al., 1993)
b1
b2
b3
b4
p1
p2
p3
p4
1 (FR84)
2 (MS90)
3 (HT91)
4 (GM79)
5 (DS88)
6 (BK67)
7 (CJ85)
8 (OY88)
0.29
0.65
0.27
0.73
0.22
0.32
0.47
-0.06
0.55
0.29
0.51
0.40
0.73
0.55
0.29
0.26
-0.10
-0.30
-0.10
-0.23
-0.05
0.00
-0.09
0.08
-0.14
-0.21
-0.24
-0.24
-0.35
0.00
-0.12
-0.03
-0.77
-0.60
-0.75
-0.68
-0.86
-0.63
-0.70
-1.00
0.10
0.10
0.13
0.13
0.26
0.05
0.13
0.31
0.27
0.27
0.12
0.24
0.34
0.00
0.28
0.25
0.14
-0.06
0.02
-0.07
-0.07
0.00
-0.04
-0.26
DR
AF
T
Model
Table B.9: Overview of the coefficients used in the various regression models, continued
(Soulsby et al., 1993)
Deltares
Model
q1
q2
q3
q4
J
1 (FR84)
2 (MS90)
3 (HT91)
4 (GM79)
5 (DS88)
6 (BK67)
7 (CJ85)
8 (OY88)
0.91
1.19
0.89
1.04
-0.89
1.14
1.65
0.38
0.25
-0.68
0.40
-0.56
2.33
0.18
-1.19
1.19
0.50
0.22
0.50
0.34
2.60
0.00
-0.42
0.25
0.45
-0.21
-0.28
-0.27
-2.50
0.00
0.49
-0.66
3.0
0.50
2.7
0.50
2.7
3.0
0.60
1.50
137
SOBEK 3, User Manual
and p and q are determined using similar equations. In this formula φ equals the angle
between the wave angle and the current angle, and the coefficients are determined by the
model index modind and tables B.8 and B.9 (related to Soulsby (1997, Table 9, p.91)):
Using the shear stresses given above, the following two Shields parameters are computed:
θm =
τm
τw
and θw =
ρg∆D50
ρg∆D50
(B.98)
Furthermore, D∗ is defined as (Soulsby, 1997, p.104):
g∆
ν2
1/3
D50
(B.99)
T
D∗ =
with which a critical Shields parameter is computed (Soulsby, 1997, eq.77, p.106):
0.30
+ 0.055 (1 − exp (−0.02D∗ ))
1 + 1.2D∗
DR
AF
θcr =
(B.100)
The sediment transport rates are computed using the following formulations for normalised
transport in current direction and normal direction (Soulsby, 1997, eq.129, p.166/167):
Φx1 = 12 (θm − θcr )
Φx2
p
θm + ε
p
= 12 (0.95 + 0.19 cos (2φ)) θm θw + ε
Φx = max (Φx1 , Φx2 )
Φy =
12
2
0.19θm θw
1.5
(θw + ε)
sin (2φ)
+ 1.5 (θm + ε)1.5
(B.101)
(B.102)
(B.103)
(B.104)
where ε is a small constant (10−4 ) to prevent numerical complications. From these expression
are finally the actual bedload transport rates obtained:
p
3
g∆D50
(Φx u − Φy v)
Sb,x =
p U
3
g∆D50
Sb,y =
(Φx v − Φy u)
U
(B.105)
(B.106)
The transport vector is imposed as bedload transport due to currents. The following formula
specific parameters have to be specified in the input files of the Transport module (see Section
B.1.3): calibration coefficient Acal , the model index for the interaction of wave and current
forces modind (integer number 1 to 8) and the D50 /z0 ratio χ (about 12).
B.4.9
Ashida–Michiue (1974)
The transport rate is given by a generalised version of the Ashida-Michiue formulation:
q
θc p
m
3
Sbc = α ∆gD50 θ
1−ξ
θ
r
1−
θc
ξ
θ
!q
(B.107)
where ξ is the hiding and exposure factor for the sediment fraction considered and:
θ=
138
q 2
C
1
∆D50
(B.108)
Deltares
Appendix: Morphology and Sediment Transport
in which q is the magnitude of the flow velocity. The transport rate is imposed as bedload
transport due to currents Sbc . The following formula specific parameters have to be specified
in the input files of the Transport module (see Section B.1.3): α, θc , m, p and q .
B.4.10
Wilcock–Crowe (2003)
The Wilcock-Crowe transport model is a fractional surface based transport model for calculating bedload transport of mixed sand and gravel sediment. The equations and their development are described in Wilcock and Crowe (2003). The bedload transport rate of each size
fraction is given by:
Wi∗ Fi U∗3
∆g

0.002φ7.5
for φ < 1.35
∗
4.5
Wi =
14 1 − 0.894
for φ ≥ 1.35
φ0.5
(B.109)
T
Sbi =
DR
AF
τ
τri
Di b
τri
=
τrm
Dm
τrm = (0.021 + 0.015 exp (−20Fs )) (ρs − ρw ) gDg
0.67
b=
Di
1 + exp 1.5 − Dg
φ=
(B.110)
(B.111)
(B.112)
(B.113)
(B.114)
where:
Di
Dg
Fi
Fs
Sbi
Wi∗
∆
τri
τrm
D50 of size fraction i
geometric mean grain size of whole grain size distribution
proportion of size fraction i on the bed surface
proportion of sand on the bed surface
bedload transort rate of size fraction i
dimensionless bedload transport rate of size fraction i
the relative density of the sediment (ρs − ρw ) /ρw
reference shear stress of grains of size Di
reference shear stress of grains of size Dg
Remarks:
The Wilcock-Crowe model incorporates its own hiding function so no external formulation should be applied.
The roughness height used for the calculation of grain shear stress during the development of the Wilcock-Crowe transport model was ks = 2D65 .
This sediment transport formula does not have any input parameters that can be, or
need to be, tuned.
B.4.11
Gaeuman et al. (2009) laboratory calibration
The Gaeuman et al. sediment transport model is a modified form of the Wilcock-Crowe model
which uses the variance of grain size distribution on the phi scale (σφ2 ) rather than the fraction of sand on the bed surface (Fs ) as a measure of the bed surface condition for use in
Deltares
139
SOBEK 3, User Manual
the calculation of reference shear stress. The ’laboratory calibration’ implementation of the
Gaeuman et al. transport model is calibrated to the experimental data used in the derivation
of the Wilcock-Crowe transport model. The model, it’s derivation and calibration is described
in Gaeuman et al. (2009).
The formulae for the calculation of Sbi , Wi∗ , φ and τri are the same as for the Wilcock-Crowe
transport model (Equations B.109, B.110, B.111 and B.112) but the calculation of τrm and b
differs.


τrm = θc0 +
0.015
 (ρs − ρw ) gDg
1 + exp 10.1σφ2 − 14.14
1 − α0
Di
1 + exp 1.5 − D
g
2
n X
Di
2
Fi
σφ2 =
log
Dg
b=
(B.115)
T
(B.116)
DR
AF
i=1
(B.117)
where θc0 and α0 are user specified parameters (See Section B.1.3). If the values θc0 = 0.021
and α0 = 0.33 are specified the original relation calibrated to the Wilcock-Crowe laboratory
data is recovered.
Remark:
The Gaeuman et al. model incorporates its own hiding function so no external formulation should be applied.
B.4.12
Gaeuman et al. (2009) Trinity River calibration
The ’Trinity River calibration’ implementation of the Gaeuman et al. transport model is calibrated to observed bedload transport rates in the Trinity River, USA and is described in Gaeuman et al. (2009). It differs from the ’laboratory calibration’ implementation in the calculation
of τrm and b.

τrm = θc0 +
b=

0.022
1 + exp
1 − α0
1 + exp 1.9 −
7.1σφ2
Di
3Dg
− 11.786
 (ρs − ρw ) gDg
(B.118)
(B.119)
where θc0 and α0 are user specified parameters (see Section B.1.3). If the values θc0 = 0.03
and α0 = 0.3 are specified the original Gaeuman et al. formulation calibrated to the Trinity
River is recovered.
Remark:
The Gaeuman et al. model incorporates its own hiding function so no external formulation should be applied.
140
Deltares
Appendix: Morphology and Sediment Transport
Morphological updating
The elevation of the bed is dynamically updated at each computational time-step. This is
one of the distinct advantages over an offline morphological computation as it means that the
hydrodynamic flow calculations are always carried out using the correct bathymetry.
At each time-step, the change in the mass of bed material that has occurred as a result of the
sediment sink and source terms and transport gradients is calculated. This change in mass is
then translated into a bed level change based on the dry bed densities of the various sediment
fractions. Both the bed levels at the cell centres and cell interfaces are updated.
T
Remark:
The depths stored at the depth points (which are read directly from the bathymetry
specified as input) are only updated for writing to the communication file and the result
files.
A number of additional features have been included in the morphological updating routine in
order to increase the flexibility. These are discussed below.
DR
AF
B.5
Morphological “switch”
You can specify whether or not to update the calculated depths to the bed by setting the
BedUpd flag in the morphology input file. It may be useful to turn bottom updating off if only
the initial patterns of erosion and deposition are required, or an investigation of sediment
transport patterns with a constant bathymetry is required.
Remark:
The use of BedUpd only affects the updating of the depth values (at ζ and velocity
points); the amount of sediment available in the bed will still be updated. If you wish
to prevent any change in both the bottom sediments and flow depths from the initial
condition then this may also be achieved by either setting the morphological delay interval MorStt to a value larger than the simulation period, or by setting the morphological
factor MorFac to 0. See below for a description of these two user variables.
Morphological delay
Frequently, a hydrodynamic simulation will take some time to stabilise after transitioning from
the initial conditions to the (dynamic) boundary conditions. It is likely that during this stabilisation period the patterns of erosion and accretion that take place do not accurately reflect
the true morphological development and should be ignored. This is made possible by use of
MorStt whereby you can specify a time interval (in minutes after the start time) after which the
morphological bottom updating will begin. During the MorStt time interval all other calculations will proceed as normal (sediment will be available for suspension for example) however
the effect of the sediment fluxes on the available bottom sediments will not be taken into
account.
Deltares
141
SOBEK 3, User Manual
Morphological time scale factor
One of the complications inherent in carrying out morphological projections on the basis of
hydrodynamic flows is that morphological developments take place on a time scale several
times longer than typical flow changes (for example, tidal flows change significantly in a period
of hours, whereas the morphology of a coastline will usually take weeks, months, or years to
change significantly). One technique for approaching this problem is to use a “morphological
time scale factor” whereby the speed of the changes in the morphology is scaled up to a rate
that it begins to have a significant impact on the hydrodynamic flows. This can be achieved
by specifying a non-unity value for the variable MorFac in the morphology input file.
T
The implementation of the morphological time scale factor is achieved by simply multiplying
the erosion and deposition fluxes from the bed to the flow and vice-versa by the MorFacfactor, at each computational time-step. This allows accelerated bed-level changes to be
incorporated dynamically into the hydrodynamic flow calculations.
DR
AF
While the maximum morphological time scale factor that can be included in a morphodynamic
model without affecting the accuracy of the model will depend on the particular situation being
modelled, and will remain a matter of judgement, tests have shown that the computations
remain stable in moderately morphologically active situations even with MorFac-factors in excess of 1 000. We also note that setting MorFac=0 is often a convenient method of preventing
both the flow depth and the quantity of sediment available at the bottom from updating, if an
investigation of a steady state solution is required.
Remarks:
Verify that the morphological factor that you use in your simulation is appropriate by
varying it (e.g. reducing it by a factor of 2) and verify that such changes do not affect
the overall simulation results.
The interpretation of the morphological factor differs for coastal and river applications.
For coastal applications with tidal motion, the morphological variations during a tidal cycle are often small and the hydrodynamics is not significantly affected by the bed level
changes. By increasing the morphological factor to for instance 10, the morphological
changes during one simulated tidal cycle are increased by this factor. From a hydrodynamical point of view this increase in morphological development rate is allowed if the
hydrodynamics is not significantly influenced. In that case the morphological development after one tidal cycle can be assumed to represent the morphological development
that would in real life only have occurred after 10 tidal cycles. In this example the number of hydrodynamic time steps required to simulate a certain period is reduced by a
factor of 10 compared to a full 1:1 simulation. This leads to a significant reduction in
simulation time. However, one should note that by following this approach the order
of events is changed, possible conflicts may arise in combination with limited sediment
availability and bed stratigraphy simulations. In river applications there is no such periodicity as a tidal cycle. For such applications, the morphological factor should be
interpreted as a speed-up factor for morphological development without changing the
order of events. Effectively, it means that the morphological development is simulated
using a, for instance 10 times, larger time step than the hydrodynamics, or phrased
more correctly the hydrodynamics is simulated at a 10 times faster rate. This means
that in case of time-varying boundary conditions (e.g. river hydrograph) the time-scale
of these forcings should be sped up: a 20 day flood peak will be compressed in 2 days.
However, one should take care that by speeding up the hydrodynamic forcings one
does not substantially change the nature of the overall hydrodynamic and morphologi-
142
Deltares
Appendix: Morphology and Sediment Transport
DR
AF
T
cal development: a quasi-steady flood period should not become a short, dynamic flash
flood. For river applications, changing the morphological factor must be associated with
changing all external time-varying forcings. For coastal applications only the overall
simulation time should be adjusted. Note that the combination of a river-like flood peak
and a tidal motion will cause problems when interpreting morphological factor not equal
to 1.
The effect of the morphological factor is different for bed and suspended load. At each
time step bedload is picked-up from the bed and deposited on the bed: only the transports are increased by the morphological factor used for the time step considered. However, in case of suspended load there is a time-delay between the time of erosion and
the time of deposition. The erosion and deposition fluxes are increased by the morphological factor, but the suspended concentrations are not (since that would influence the
density effects). It is possible to vary the morphological factor during a simulation to
speed up relatively quiet periods more than relatively active periods. Such changes in
the morphological factor will not influence the mass balance of a bed or total load simulation since pickup and deposition are combined into one time step. However, in case
of suspended load the entrainment and deposition may occur at time-steps governed
by different morphological factors. In such cases the entrainment flux that generated a
certain suspended sediment concentration will differ from the deposition flux that was
caused by the settling of the same suspended sediment. A change in morphological
factor during a period of non-zero suspended sediment concentrations, will thus lead to
a mass-balance error in the order of the suspended sediment volume times the change
in morphological factor. The error may kept to a minimum by appropriately choosing the
transition times.
Deltares
143
DR
AF
T
T
DR
AF
Photo’s by: BeeldbankVenW.nl, Rijkswaterstaat / Joop van Houdt.
T+31 (0)88 335 81 88
F +31 (0)88 335 81 11
[email protected]
www.deltaressystems.nl